Massive Online

Analysis
Manual
Albert Bifet and Richard Kirkby

August 2009
 Contents

1 Introduction 1
1.1 Data streams Evaluation . . . . . . . . . . . . . . . . . . . . . 2

2 Installation 5

3 Using the GUI 7

4 Using the command line 9

4.1 Comparing two classifiers . . . . . . . . . . . . . . . . . . . . 11

5 Tasks in MOA 13
5.1 WriteStreamToARFFFile . . . . . . . . . . . . . . . . . . . . . 13
5.2 MeasureStreamSpeed . . . . . . . . . . . . . . . . . . . . . . . 13
5.3 LearnModel . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.4 EvaluateModel . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.5 EvaluatePeriodicHeldOutTest . . . . . . . . . . . . . . . . . . 15
5.6 EvaluateInterleavedTestThenTrain . . . . . . . . . . . . . . . 15
5.7 EvaluatePrequential . . . . . . . . . . . . . . . . . . . . . . . 16

6 Evolving data streams 19

6.1 Streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
6.1.1 ArffFileStream . . . . . . . . . . . . . . . . . . . . . . 19
6.1.2 ConceptDriftStream . . . . . . . . . . . . . . . . . . . 20
6.1.3 ConceptDriftRealStream . . . . . . . . . . . . . . . . . 21
6.1.4 FilteredStream . . . . . . . . . . . . . . . . . . . . . . 22
6.1.5 AddNoiseFilter . . . . . . . . . . . . . . . . . . . . . . 22
6.2 Streams Generators . . . . . . . . . . . . . . . . . . . . . . . . 22
6.2.1 generators.AgrawalGenerator . . . . . . . . . . . . . 22
6.2.2 generators.HyperplaneGenerator . . . . . . . . . . . . 23
6.2.3 generators.LEDGenerator . . . . . . . . . . . . . . . . 24
6.2.4 generators.LEDGeneratorDrift . . . . . . . . . . . . . 24
6.2.5 generators.RandomRBFGenerator . . . . . . . . . . . 25
6.2.6 generators.RandomRBFGeneratorDrift . . . . . . . . 25
6.2.7 generators.RandomTreeGenerator . . . . . . . . . . . 26
6.2.8 generators.SEAGenerator . . . . . . . . . . . . . . . . 27
6.2.9 generators.STAGGERGenerator . . . . . . . . . . . . 27
6.2.10 generators.WaveformGenerator . . . . . . . . . . . . 28
6.2.11 generators.WaveformGeneratorDrift . . . . . . . . . . 28

i
CONTENTS

7 Classifiers 29
7.1 Classifiers for static streams . . . . . . . . . . . . . . . . . . . 29
7.1.1 MajorityClass . . . . . . . . . . . . . . . . . . . . . . . 29
7.1.2 Naive Bayes . . . . . . . . . . . . . . . . . . . . . . . . 30
7.1.3 DecisionStump . . . . . . . . . . . . . . . . . . . . . . 30
7.1.4 HoeffdingTree . . . . . . . . . . . . . . . . . . . . . . . 31
7.1.5 HoeffdingTreeNB . . . . . . . . . . . . . . . . . . . . . 32
7.1.6 HoeffdingTreeNBAdaptive . . . . . . . . . . . . . . . 32
7.1.7 HoeffdingOptionTree . . . . . . . . . . . . . . . . . . 33
7.1.8 HoeffdingOptionTreeNB . . . . . . . . . . . . . . . . . 34
7.1.9 HoeffdingTreeOptionNBAdaptive . . . . . . . . . . . 34
7.1.10 OzaBag . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
7.1.11 OzaBoost . . . . . . . . . . . . . . . . . . . . . . . . . . 35
7.1.12 OCBoost . . . . . . . . . . . . . . . . . . . . . . . . . . 35
7.2 Classifiers for evolving streams . . . . . . . . . . . . . . . . . 36
7.2.1 OzaBagASHT . . . . . . . . . . . . . . . . . . . . . . . 36
7.2.2 OzaBagADWIN . . . . . . . . . . . . . . . . . . . . . . 37
7.2.3 SingleClassifierDrift . . . . . . . . . . . . . . . . . . . 39
7.2.4 AdaHoeffdingOptionTree . . . . . . . . . . . . . . . . 40

8 Writing a classifier 41
8.1 Creating a new classifier . . . . . . . . . . . . . . . . . . . . . 41
8.2 Compiling a classifier . . . . . . . . . . . . . . . . . . . . . . . 46

9 Bi-directional interface with WEKA 47

9.1 WEKA classifiers from MOA . . . . . . . . . . . . . . . . . . 47
9.1.1 WekaClassifier . . . . . . . . . . . . . . . . . . . . . . 48
9.1.2 SingleClassifierDrift . . . . . . . . . . . . . . . . . . . 48
9.2 MOA classifiers from WEKA . . . . . . . . . . . . . . . . . . 50

ii
 Introduction
1
Massive Online Analysis (MOA) is a software environment for imple-
menting algorithms and running experiments for online learning from
evolving data streams.

MOA includes a collection of offline and online methods as well as

tools for evaluation. In particular, it implements boosting, bagging, and
Hoeffding Trees, all with and without Naı̈ve Bayes classifiers at the leaves.
MOA is related to WEKA, the Waikato Environment for Knowledge
Analysis, which is an award-winning open-source workbench contain-
ing implementations of a wide range of batch machine learning methods.
WEKA is also written in Java. The main benefits of Java are portability,
where applications can be run on any platform with an appropriate Java
virtual machine, and the strong and well-developed support libraries. Use
of the language is widespread, and features such as the automatic garbage
collection help to reduce programmer burden and error.
One of the key data structures used in MOA is the description of an
example from a data stream. This structure borrows from WEKA, where
an example is represented by an array of double precision floating point
values. This provides freedom to store all necessary types of value – nu-
meric attribute values can be stored directly, and discrete attribute values
and class labels are represented by integer index values that are stored as
floating point values in the array. Double precision floating point values
require storage space of 64 bits, or 8 bytes. This detail can have implica-
tions for memory usage.

1
CHAPTER 1. INTRODUCTION

Figure 1.1: The data stream classification cycle

1.1 Data streams Evaluation

A data stream environment has different requirements from the traditional
setting. The most significant are the following:

Requirement 1 Process an example at a time, and inspect it only once (at

most)

Requirement 2 Use a limited amount of memory

Requirement 3 Work in a limited amount of time

Requirement 4 Be ready to predict at any time

We have to consider these requirements in order to design a new experi-

mental framework for data streams. Figure 1.1 illustrates the typical use
of a data stream classification algorithm, and how the requirements fit in
a repeating cycle:

1. The algorithm is passed the next available example from the stream
(requirement 1).

2
 1.1. DATA STREAMS EVALUATION

2. The algorithm processes the example, updating its data structures.

It does so without exceeding the memory bounds set on it (require-
ment 2), and as quickly as possible (requirement 3).
3. The algorithm is ready to accept the next example. On request it is
able to predict the class of unseen examples (requirement 4).
In traditional batch learning the problem of limited data is overcome by
analyzing and averaging multiple models produced with different ran-
dom arrangements of training and test data. In the stream setting the
problem of (effectively) unlimited data poses different challenges. One
solution involves taking snapshots at different times during the induction
of a model to see how much the model improves.
The evaluation procedure of a learning algorithm determines which ex-
amples are used for training the algorithm, and which are used to test the
model output by the algorithm. The procedure used historically in batch
learning has partly depended on data size. As data sizes increase, practical
time limitations prevent procedures that repeat training too many times. It
is commonly accepted with considerably larger data sources that it is nec-
essary to reduce the numbers of repetitions or folds to allow experiments
to complete in reasonable time. When considering what procedure to use
in the data stream setting, one of the unique concerns is how to build a
picture of accuracy over time. Two main approaches arise:
• Holdout: When traditional batch learning reaches a scale where cross-
validation is too time consuming, it is often accepted to instead mea-
sure performance on a single holdout set. This is most useful when
the division between train and test sets have been pre-defined, so
that results from different studies can be directly compared.
• Interleaved Test-Then-Train or Prequential: Each individual exam-
ple can be used to test the model before it is used for training, and
from this the accuracy can be incrementally updated. When inten-
tionally performed in this order, the model is always being tested
on examples it has not seen. This scheme has the advantage that no
holdout set is needed for testing, making maximum use of the avail-
able data. It also ensures a smooth plot of accuracy over time, as
each individual example will become increasingly less significant to
the overall average.
As data stream classification is a relatively new field, such evaluation prac-
tices are not nearly as well researched and established as they are in the
traditional batch setting. The majority of experimental evaluations use less

3
CHAPTER 1. INTRODUCTION

than one million training examples. In the context of data streams this is
disappointing, because to be truly useful at data stream classification the
algorithms need to be capable of handling very large (potentially infinite)
streams of examples. Demonstrating systems only on small amounts of
data does not build a convincing case for capacity to solve more demand-
ing data stream applications.
MOA permits adequately evaluate data stream classification algorithms
on large streams, in the order of tens of millions of examples where possi-
ble, and under explicit memory limits. Any less than this does not actually
test algorithms in a realistically challenging setting.

4
 Installation
2
The following manual is based on a Unix/Linux system with Java 5 SDK
or greater installed. Other operating systems such as Microsoft Windows
will be similar but may need adjustments to suit.
MOA needs the following files:
moa.jar
weka.jar
sizeofag.jar
They are available from
http://sourceforge.net/projects/moa-datastream/
http://sourceforge.net/projects/weka/
http://www.jroller.com/resources/m/maxim/sizeofag.jar
These files are needed to run the MOA software from the command
line and the graphical interface:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
LearnModel -l DecisionStumpTutorial \
-s generators.WaveformGenerator -m 1000000 -O model1.moa

java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar \

moa.gui.TaskLauncher
or, using Microsoft Windows:
java -cp .;moa.jar;weka.jar -javaagent:sizeofag.jar moa.DoTask
LearnModel -l DecisionStumpTutorial
-s generators.WaveformGenerator -m 1000000 -O model1.moa

java -cp .;moa.jar;weka.jar -javaagent:sizeofag.jar

moa.gui.TaskLauncher

5
 Using the GUI
3
A graphical user interface for configuring and running tasks is available
with the command:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar \
moa.gui.TaskLauncher
Click ’Configure’ to set up a task, when ready click to launch a task
click ’Run’. Several tasks can be run concurrently. Click on different tasks
in the list and control them using the buttons below. If textual output of a
task is available it will be displayed in the bottom half of the GUI, and can
be saved to disk.

Figure 3.1: Graphical user interface of MOA

Note that the command line text box displayed at the top of the win-
dow represents textual commands that can be used to run tasks on the
command line as described in the next chapter. The text can be selected
then copied onto the clipboard.

7
CHAPTER 3. USING THE GUI

Figure 3.2: Options to set up a task in MOA

8
 Using the command line
4
In this chapter we are going to show some examples of tasks performed
using the command line.
The first example will command MOA to train the HoeffdingTree
classifier and create a model. The moa.DoTask class is the main class for
running tasks on the command line. It will accept the name of a task fol-
lowed by any appropriate parameters. The first task used is the LearnModel
task. The -l parameter specifies the learner, in this case the HoeffdingTree
class. The -s parameter specifies the stream to learn from, in this case
generators.WaveformGenerator is specified, which is a data stream
generator that produces a three-class learning problem of identifying three
types of waveform. The -m option specifies the maximum number of ex-
amples to train the learner with, in this case one million examples. The -O
option specifies a file to output the resulting model to:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
LearnModel -l HoeffdingTree \
-s generators.WaveformGenerator -m 1000000 -O model1.moa

This will create a file named model1.moa that contains the decision
stump model that was induced during training.
The next example will evaluate the model to see how accurate it is on
a set of examples that are generated using a different random seed. The
EvaluateModel task is given the parameters needed to load the model
produced in the previous step, generate a new waveform stream with a
random seed of 2, and test on one million examples:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
"EvaluateModel -m file:model1.moa \
-s (generators.WaveformGenerator -i 2) -i 1000000"

This is the first example of nesting parameters using brackets. Quotes

have been added around the description of the task, otherwise the operat-
ing system may be confused about the meaning of the brackets.

9
CHAPTER 4. USING THE COMMAND LINE

After evaluation the following statistics are output:

classified instances = 1,000,000
classifications correct (percent) = 57.637

Note the the above two steps can be achieved by rolling them into one,
avoiding the need to create an external file, as follows:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
"EvaluateModel -m (LearnModel -l HoeffdingTree \
-s generators.WaveformGenerator -m 1000000) \
-s (generators.WaveformGenerator -i 2) -i 1000000"

The task EvaluatePeriodicHeldOutTest will train a model while

taking snapshots of performance using a held-out test set at periodic inter-
vals. The following command creates a comma separated values file, training
the HoeffdingTree classifier on the WaveformGenerator data, using
the first 100 thousand examples for testing, training on a total of 100 mil-
lion examples, and testing every one million examples:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
"EvaluatePeriodicHeldOutTest -l HoeffdingTree \
-s generators.WaveformGenerator \
-n 100000 -i 100000000 -f 1000000" > dsresult.csv

For the purposes of comparison, a bagging learner using ten decisions

trees can be trained on the same problem:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
"EvaluatePeriodicHeldOutTest -l (OzaBag -l HoeffdingTree -s 10)\
-s generators.WaveformGenerator \
-n 100000 -i 100000000 -f 1000000" > htresult.csv

Another evaluation method implemented in MOA is Interleaved Test-

Then-Train or Prequential: Each individual example is used to test the model
before it is used for training, and from this the accuracy is incrementally
updated. When intentionally performed in this order, the model is always
being tested on examples it has not seen. This scheme has the advantage
that no holdout set is needed for testing, making maximum use of the
available data. It also ensures a smooth plot of accuracy over time, as each
individual example will become increasingly less significant to the overall
average.
An example of the EvaluateInterleavedTestThenTrain task creating a
comma separated values file, training the HoeffdingTree classifier on the
WaveformGenerator data, training and testing on a total of 100 million
examples, and testing every one million examples, is the following:

10
 4.1. COMPARING TWO CLASSIFIERS

java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \

"EvaluateInterleavedTestThenTrain -l HoeffdingTree \
-s generators.WaveformGenerator \
-i 100000000 -f 1000000" > htresult.csv

4.1 Comparing two classifiers

Suppose we would like to compare the learning curves of a decision stump
and a Hoeffding Tree. First, we will execute the task EvaluatePeriodicHeldOutTest
to train a model while taking snapshots of performance using a held-out
test set at periodic intervals. The following commands create comma sepa-
rated values files, training the DecisionStump and the HoeffdingTree
classifier on the WaveformGenerator data, using the first 100 thousand
examples for testing, training on a total of 100 million examples, and test-
ing every one million examples:

java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \

"EvaluatePeriodicHeldOutTest -l DecisionStump \
-s generators.WaveformGenerator \
-n 100000 -i 100000000 -f 1000000" > dsresult.csv

java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \

"EvaluatePeriodicHeldOutTest -l HoeffdingTree \
-s generators.WaveformGenerator \
-n 100000 -i 100000000 -f 1000000" > htresult.csv

Assuming that gnuplot is installed on the system, the learning curves

can be plotted with the following commands:

gnuplot> set datafile separator ","

gnuplot> set ylabel "% correct"
gnuplot> set xlabel "examples processed"
gnuplot> plot [][0:100] \
"dsresult.csv" using 1:9 with linespoints \
title "DecisionStumpTutorial", \
"htresult.csv" using 1:9 with linespoints \
title "HoeffdingTree"

This results in the following graph:

11
CHAPTER 4. USING THE COMMAND LINE

100
DecisionStumpTutorial
HoeffdingTree

80

60
% correct

40

20

0
1e+06 2e+06 3e+06 4e+06 5e+06 6e+06 7e+06 8e+06 9e+06 1e+07
examples processed
For this problem it is obvious that a full tree can achieve higher accu-
racy than a single stump, and that a stump has very stable accuracy that
does not improve with further training.

12
 5
Tasks in MOA

The main Tasks in MOA are the following:

5.1 WriteStreamToARFFFile
Outputs a stream to an ARFF file. Example:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
"WriteStreamToARFFFile -s generators.WaveformGenerator \
-f Wave.arff -m 100000"

Parameters:
• -s : Stream to write
• -f : Destination ARFF file
• -m : Maximum number of instances to write to file
• -h : Suppress header from output

5.2 MeasureStreamSpeed
Measures the speed of a stream. Example:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
"MeasureStreamSpeed -s generators.WaveformGenerator \
-g 100000"

Parameters:
• -s : Stream to measure
• -g : Number of examples
• -O : File to save the final result of the task to

13
CHAPTER 5. TASKS IN MOA

5.3 LearnModel
Learns a model from a stream. Example:

java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \

LearnModel -l HoeffdingTree \
-s generators.WaveformGenerator -m 1000000 -O model1.moa

Parameters:

• -l : Classifier to train

• -s : Stream to learn from

• -m : Maximum number of instances to train on per pass over the data

• -p : The number of passes to do over the data

• -b : Maximum size of model (in bytes). -1 = no limit

• -q : How many instances between memory bound checks

• -O : File to save the final result of the task to

5.4 EvaluateModel
Evaluates a static model on a stream. Example:

java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \

"EvaluateModel -m (LearnModel -l HoeffdingTree \
-s generators.WaveformGenerator -m 1000000) \
-s (generators.WaveformGenerator -i 2) -i 1000000"

Parameters:

• -l : Classifier to evaluate

• -s : Stream to evaluate on

• -e : Classification performance evaluation method

• -i : Maximum number of instances to test

• -O : File to save the final result of the task to

14
 5.5. EVALUATEPERIODICHELDOUTTEST

5.5 EvaluatePeriodicHeldOutTest
Evaluates a classifier on a stream by periodically testing on a heldout set.
Example:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
"EvaluatePeriodicHeldOutTest -l HoeffdingTree \
-s generators.WaveformGenerator \
-n 100000 -i 100000000 -f 1000000" > htresult.csv

Parameters:

• -l : Classifier to train

• -s : Stream to learn from

• -e : Classification performance evaluation method

• -n : Number of testing examples

• -i : Number of training examples, <1 = unlimited

• -t : Number of training seconds

• -f : Number of training examples between samples of learning per-

formance

• -d : File to append intermediate csv results to

• -c : Cache test instances in memory

• -O : File to save the final result of the task to

5.6 EvaluateInterleavedTestThenTrain
Evaluates a classifier on a stream by testing then training with each exam-
ple in sequence. Example:
java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \
"EvaluateInterleavedTestThenTrain -l HoeffdingTree \
-s generators.WaveformGenerator \
-i 100000000 -f 1000000" > htresult.csv

Parameters:

• -l : Classifier to train

15
CHAPTER 5. TASKS IN MOA

• -s : Stream to learn from

• -e : Classification performance evaluation method

• -n : Number of testing examples

• -i : Maximum number of instances to test/train on (-1 = no limit)

• -t : Maximum number of seconds to test/train for (-1 = no limit)

• -f : How many instances between samples of the learning perfor-

mance

• -b : Maximum size of model (in bytes). -1 = no limit

• -q : How many instances between memory bound checks

• -d : File to append intermediate csv results to

• -O : File to save the final result of the task to

5.7 EvaluatePrequential
Evaluates a classifier on a stream by testing then training with each exam-
ple in sequence. It may use a sliding window or a fading factor forgetting
mechanism.
This evaluation method using sliding windows and a fading factor was
presented in

[C] João Gama, Raquel Sebastião and Pedro Pereira Rodrigues. Issues in
evaluation of stream learning algorithms. In KDD’09, pages 329–338.

The fading factor α is used as following:

Si
Ei =
Bi
with
Si = Li + α × Si−1
Bi = ni + α × Bi−1
where ni is the number of examples used to compute the loss function Li .
ni = 1 since the loss Li is computed for every single example.
Examples:

16
 5.7. EVALUATEPREQUENTIAL

java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \

"EvaluatePrequential -l HoeffdingTree \
-e WindowClassificationPerformanceEvaluator -w 10000 \
-s generators.WaveformGenerator \
-i 100000000 -f 1000000" > htresult.csv

java -cp .:moa.jar:weka.jar -javaagent:sizeofag.jar moa.DoTask \

"EvaluatePrequential -l HoeffdingTree \
-e FadingFactorClassificationPerformanceEvaluator -a .975 \
-s generators.WaveformGenerator \
-i 100000000 -f 1000000" > htresult.csv

Parameters:

• Same parameters as EvaluateInterleavedTestThenTrain

• -e : Classification performance evaluation method

– WindowClassificationPerformanceEvaluator
– FadingFactorClassificationPerformanceEvaluator
– EWMAFactorClassificationPerformanceEvaluator

• -w : Size of sliding window to use with WindowClassificationPerfor-

manceEvaluator

• -a : Fading factor to use with FadingFactorClassificationPerformanceE-

valuator

17
 Evolving data streams
6
MOA streams are build using generators, reading ARFF files, joining sev-
eral streams, or filtering streams. MOA streams generators allow to simu-
late potentially infinite sequence of data. There are the following :
• Random Tree Generator
• SEA Concepts Generator
• STAGGER Concepts Generator
• Rotating Hyperplane
• Random RBF Generator
• LED Generator
• Waveform Generator
• Function Generator
.

6.1 Streams
Classes available in MOA to obtain input streams are the following:

6.1.1 ArffFileStream
A stream read from an ARFF file. Example:
ArffFileStream -f elec.arff
Parameters:
• -f : ARFF file to load
• -c : Class index of data. 0 for none or -1 for last attribute in file

19
CHAPTER 6. EVOLVING DATA STREAMS

f(t)
1 f(t)

α
0.5

α
t
t0
W

Figure 6.1: A sigmoid function f(t) = 1/(1 + e−s(t−t0 ) ).

6.1.2 ConceptDriftStream
Generator that adds concept drift to examples in a stream.
Considering data streams as data generated from pure distributions,
MOA models a concept drift event as a weighted combination of two pure
distributions that characterizes the target concepts before and after the
drift. MOA uses the sigmoid function, as an elegant and practical solution
to define the probability that every new instance of the stream belongs to
the new concept after the drift.
We see from Figure 6.1 that the sigmoid function
f(t) = 1/(1 + e−s(t−t0 ) )
has a derivative at the point t0 equal to f 0 (t0 ) = s/4. The tangent of angle
α is equal to this derivative, tan α = s/4. We observe that tan α = 1/W,
and as s = 4 tan α then s = 4/W. So the parameter s in the sigmoid gives
the length of W and the angle α. In this sigmoid model we only need
to specify two parameters : t0 the point of change, and W the length of
change. Note that for any positive real number β
f(t0 + β · W) = 1 − f(t0 − β · W),
and that f(t0 +β·W) and f(t0 −β·W) are constant values that don’t depend
on t0 and W:
f(t0 + W/2) = 1 − f(t0 − W/2) = 1/(1 + e−2 ) ≈ 88.08%
f(t0 + W) = 1 − f(t0 − W) = 1/(1 + e−4 ) ≈ 98.20%
f(t0 + 2W) = 1 − f(t0 − 2W) = 1/(1 + e−8 ) ≈ 99.97%

20
 6.1. STREAMS

Definition 1. Given two data streams a, b, we define c = a ⊕W t0 b as the data

stream built joining the two data streams a and b, where t0 is the point of change,
W is the length of change and

• Pr[c(t) = a(t)] = e−4(t−t0 )/W /(1 + e−4(t−t0 )/W )

• Pr[c(t) = b(t)] = 1/(1 + e−4(t−t0 )/W ).

Example:

ConceptDriftStream -s (generators.AgrawalGenerator -f 7)
-d (generators.AgrawalGenerator -f 2) -w 1000000 -p 900000

Parameters:

• -s : Stream

• -d : Concept drift Stream

• -p : Central position of concept drift change

• -w : Width of concept drift change

6.1.3 ConceptDriftRealStream
Generator that adds concept drift to examples in a stream with different
classes and attributes. Example: real datasets
Example:

ConceptDriftRealStream -s (ArffFileStream -f covtype.arff) \

-d (ConceptDriftRealStream -s (ArffFileStream -f PokerOrig.arff) \
-d (ArffFileStream -f elec.arff) -w 5000 -p 1000000 ) -w 5000 -p 581012

Parameters:

• -s : Stream

• -d : Concept drift Stream

• -p : Central position of concept drift change

• -w : Width of concept drift change

21
CHAPTER 6. EVOLVING DATA STREAMS

6.1.4 FilteredStream
A stream that is filtered.
Parameters:

• -s : Stream to filter

• -f : Filters to apply : AddNoiseFilter

6.1.5 AddNoiseFilter
Adds random noise to examples in a stream. Only to use with Filtered-
Stream.
Parameters:

• -r : Seed for random noise

• -a : The fraction of attribute values to disturb

• -c : The fraction of class labels to disturb

6.2 Streams Generators

The classes available to generate streams are the following:

6.2.1 generators.AgrawalGenerator
Generates one of ten different pre-defined loan functions
It was introduced by Agrawal et al. in
[A] R. Agrawal, T. Imielinski, and A. Swami. Database mining: A per-
formance perspective. IEEE Trans. on Knowl. and Data Eng., 5(6):914–
925, 1993.
It was a common source of data for early work on scaling up decision
tree learners. The generator produces a stream containing nine attributes,
six numeric and three categorical. Although not explicitly stated by the
authors, a sensible conclusion is that these attributes describe hypothetical
loan applications. There are ten functions defined for generating binary
class labels from the attributes. Presumably these determine whether the
loan should be approved.
A public C source code is available. The built in functions are based on
the paper (page 924), which turn out to be functions pred20 thru pred29

22
 6.2. STREAMS GENERATORS

in the public C implementation Perturbation function works like C imple-

mentation rather than description in paper
Parameters:

• -f : Classification function used, as defined in the original paper.

• -i : Seed for random generation of instances.

• -p : The amount of peturbation (noise) introduced to numeric values

• -b : Balance the number of instances of each class.

6.2.2 generators.HyperplaneGenerator
Generates a problem of predicting class of a rotating hyperplane.
It was used as testbed for CVFDT versus VFDT in

[C] G. Hulten, L. Spencer, and P. Domingos. Mining time-changing data

streams. In KDD’01, pages 97–106, San Francisco, CA, 2001. ACM
Press.

A hyperplane in d-dimensional space is the set of points x that satisfy

X
d X
d
wi x i = w0 = wi
i=1 i=1

P
where xi , is the ith coordinate of x. Examples for which di=1 wi xi ≥ w0
P
are labeled positive, and examples for which di=1 wi xi < w0 are labeled
negative. Hyperplanes are useful for simulating time-changing concepts,
because we can change the orientation and position of the hyperplane in a
smooth manner by changing the relative size of the weights. We introduce
change to this dataset adding drift to each weight attribute wi = wi + dσ,
where σ is the probability that the direction of change is reversed and d is
the change applied to every example.
Parameters:

• -i : Seed for random generation of instances.

• -c : The number of classes to generate

• -a : The number of attributes to generate.

• -k : The number of attributes with drift.

23
CHAPTER 6. EVOLVING DATA STREAMS

• -t : Magnitude of the change for every example

• -n : Percentage of noise to add to the data.

• -s : Percentage of probability that the direction of change is reversed

6.2.3 generators.LEDGenerator
Generates a problem of predicting the digit displayed on a 7-segment LED
display.
This data source originates from the CART book. An implementation
in C was donated to the UCI machine learning repository by David Aha.
The goal is to predict the digit displayed on a seven-segment LED display,
where each attribute has a 10% chance of being inverted. It has an optimal
Bayes classification rate of 74%. The particular configuration of the gener-
ator used for experiments (led) produces 24 binary attributes, 17 of which
are irrelevant.
Parameters:

• -i : Seed for random generation of instances.

• -n : Percentage of noise to add to the data

• -s : Reduce the data to only contain 7 relevant binary attributes

6.2.4 generators.LEDGeneratorDrift
Generates a problem of predicting the digit displayed on a 7-segment LED
display with drift.
Parameters:

• -i : Seed for random generation of instances.

• -n : Percentage of noise to add to the data

• -s : Reduce the data to only contain 7 relevant binary attributes

• -d : Number of attributes with drift

24
 6.2. STREAMS GENERATORS

6.2.5 generators.RandomRBFGenerator
Generates a random radial basis function stream.
This generator was devised to offer an alternate complex concept type
that is not straightforward to approximate with a decision tree model. The
RBF (Radial Basis Function) generator works as follows: A fixed number
of random centroids are generated. Each center has a random position, a
single standard deviation, class label and weight. New examples are gen-
erated by selecting a center at random, taking weights into consideration
so that centers with higher weight are more likely to be chosen. A random
direction is chosen to offset the attribute values from the central point.
The length of the displacement is randomly drawn from a Gaussian dis-
tribution with standard deviation determined by the chosen centroid. The
chosen centroid also determines the class label of the example. This effec-
tively creates a normally distributed hypersphere of examples surround-
ing each central point with varying densities. Only numeric attributes are
generated.
Parameters:

• -r : Seed for random generation of model

• -i : Seed for random generation of instances

• -c : The number of classes to generate

• -a : The number of attributes to generate

• -n : The number of centroids in the model

6.2.6 generators.RandomRBFGeneratorDrift
Generates a random radial basis function stream with drift. Drift is intro-
duced by moving the centroids with constant speed.
Parameters:

• -r : Seed for random generation of model

• -i : Seed for random generation of instances

• -c : The number of classes to generate

• -a : The number of attributes to generate

• -n : The number of centroids in the model

25
CHAPTER 6. EVOLVING DATA STREAMS

• -s : Speed of change of centroids in the model.

• -k : The number of centroids with drift

6.2.7 generators.RandomTreeGenerator
Generates a stream based on a randomly generated tree.
This generator is based on that proposed in

[D] P. Domingos and G. Hulten. Mining high-speed data streams. In

Knowledge Discovery and Data Mining, pages 71–80, 2000.

It produces concepts that in theory should favour decision tree learn-

ers. It constructs a decision tree by choosing attributes at random to split,
and assigning a random class label to each leaf. Once the tree is built,
new examples are generated by assigning uniformly distributed random
values to attributes which then determine the class label via the tree.
The generator has parameters to control the number of classes, attributes,
nominal attribute labels, and the depth of the tree. For consistency be-
tween experiments, two random trees were generated and fixed as the
base concepts for testingone simple and the other complex, where com-
plexity refers to the number of attributes involved and the size of the tree.
A degree of noise can be introduced to the examples after generation.
In the case of discrete attributes and the class label, a probability of noise
parameter determines the chance that any particular value is switched to
something other than the original value. For numeric attributes, a degree
of random noise is added to all values, drawn from a random Gaussian
distribution with standard deviation equal to the standard deviation of
the original values multiplied by noise probability.
Parameters:

• -r: Seed for random generation of tree

• -i: Seed for random generation of instances

• -c: The number of classes to generate

• -o: The number of nominal attributes to generate

• -u: The number of numeric attributes to generate

• -v: The number of values to generate per nominal attribute

• -d: The maximum depth of the tree concept

26
 6.2. STREAMS GENERATORS

• -l: The first level of the tree above maxTreeDepth that can have leaves

• -f: The fraction of leaves per level from firstLeafLevel onwards

6.2.8 generators.SEAGenerator
Generates SEA concepts functions. This dataset contains abrupt concept
drift, first introduced in paper:

[S] W. N. Street and Y. Kim. A streaming ensemble algorithm (SEA) for

large-scale classification. In KDD ’01, pages 377–382, New York, NY,
USA, 2001. ACM Press.

It is generated using three attributes, where only the two first attributes are
relevant. All three attributes have values between 0 and 10. The points of
the dataset are divided into 4 blocks with different concepts. In each block,
the classification is done using f1 + f2 ≤ θ, where f1 and f2 represent the
first two attributes and θ is a threshold value. The most frequent values
are 9, 8, 7 and 9.5 for the data blocks.
Parameters:

• -f: Classification function used, as defined in the original paper

• -i: Seed for random generation of instances

• -b: Balance the number of instances of each class

• -n: Percentage of noise to add to the data

6.2.9 generators.STAGGERGenerator
Generates STAGGER Concept functions. They were introduced by Schlim-
mer and Granger in

[ST] J. C. Schlimmer and R. H. Granger. Incremental learning from noisy

data. Machine Learning, 1(3):317–354, 1986.

The STAGGER Concepts are boolean functions of three attributes en-

coding objects: size (small, medium, and large), shape (circle, triangle,
and rectangle), and colour (red,blue, and green). A concept description
covering either green rectangles or red triangles is represented by (shape=
rectangle and colour=green) or (shape=triangle and colour=red).
Parameters:

27
CHAPTER 6. EVOLVING DATA STREAMS

1. -i: Seed for random generation of instances

2. -f: Classification function used, as defined in the original paper

3. -b: Balance the number of instances of each class

6.2.10 generators.WaveformGenerator
Generates a problem of predicting one of three waveform types.
It shares its origins with LED, and was also donated by David Aha to
the UCI repository. The goal of the task is to differentiate between three
different classes of waveform, each of which is generated from a combi-
nation of two or three base waves. The optimal Bayes classification rate is
known to be 86%. There are two versions of the problem, wave21 which
has 21 numeric attributes, all of which include noise, and wave40 which
introduces an additional 19 irrelevant attributes.
Parameters:

• -i: Seed for random generation of instances

• -n: Adds noise, for a total of 40 attributes

6.2.11 generators.WaveformGeneratorDrift
Generates a problem of predicting one of three waveform types with drift.
Parameters:

• -i: Seed for random generation of instances

• -n: Adds noise, for a total of 40 attributes

• -d: Number of attributes with drift

28
 7
Classifiers

The classifiers implemented in MOA are the following:

• Naive Bayes

• Decision Stump

• Hoeffding Tree

• Hoeffding Option Tree

• Bagging

• Boosting

• Bagging using ADWIN

• Bagging using Adaptive-Size Hoeffding Trees.

7.1 Classifiers for static streams

7.1.1 MajorityClass
Always predicts the class that has been observed most frequently the in
the training data.
Parameters:

• -r : Seed for random behaviour of the classifier

29
CHAPTER 7. CLASSIFIERS

7.1.2 Naive Bayes

Performs classic bayesian prediction while making naive assumption that
all inputs are independent.
Naı̈ve Bayes is a classifier algorithm known for its simplicity and low
computational cost. Given nC different classes, the trained Naı̈ve Bayes
classifier predicts for every unlabelled instance I the class C to which it
belongs with high accuracy.
The model works as follows: Let x1 ,. . . , xk be k discrete attributes, and
assume that xi can take ni different values. Let C be the class attribute,
which can take nC different values. Upon receiving an unlabelled instance
I = (x1 = v1 , . . . , xk = vk ), the Naı̈ve Bayes classifier computes a “proba-
bility” of I being in class c as:

Y
k
∼
Pr[C = c|I] = Pr[xi = vi |C = c]
i=1
Yk
Pr[xi = vi ∧ C = c]
= Pr[C = c] ·
i=1
Pr[C = c]

The values Pr[xi = vj ∧ C = c] and Pr[C = c] are estimated from

the training data. Thus, the summary of the training data is simply a 3-
dimensional table that stores for each triple (xi , vj , c) a count Ni,j,c of train-
ing instances with xi = vj , together with a 1-dimensional table for the
counts of C = c. This algorithm is naturally incremental: upon receiving a
new example (or a batch of new examples), simply increment the relevant
counts. Predictions can be made at any time from the current counts.
Parameters:
• -r : Seed for random behaviour of the classifier

7.1.3 DecisionStump
Decision trees of one level.
Parameters:
• -g : The number of instances to observe between model changes
• -b : Only allow binary splits
• -c : Split criterion to use. Example : InfoGainSplitCriterion
• -r : Seed for random behaviour of the classifier

30
 7.1. CLASSIFIERS FOR STATIC STREAMS

7.1.4 HoeffdingTree
Decision tree for streaming data.
A Hoeffding tree is an incremental, anytime decision tree induction al-
gorithm that is capable of learning from massive data streams, assuming
that the distribution generating examples does not change over time. Ho-
effding trees exploit the fact that a small sample can often be enough to
choose an optimal splitting attribute. This idea is supported mathemat-
ically by the Hoeffding bound, which quantifies the number of observa-
tions (in our case, examples) needed to estimate some statistics within a
prescribed precision (in our case, the goodness of an attribute). More pre-
cisely, the Hoeffding bound states that with probability 1−δ, the true mean
of a random variable of range R will not differ from the estimated mean
after n independent observations by more than:
r
R2 ln(1/δ)
= .
2n
A theoretically appealing feature of Hoeffding Trees not shared by other
incremental decision tree learners is that it has sound guarantees of per-
formance. Using the Hoeffding bound one can show that its output is
asymptotically nearly identical to that of a non-incremental learner using
infinitely many examples. See for details:

[C] G. Hulten, L. Spencer, and P. Domingos. Mining time-changing data

streams. In KDD’01, pages 97–106, San Francisco, CA, 2001. ACM
Press.

Parameters:

• -m : Maximum memory consumed by the tree

• -n : Numeric estimator to use :

– Gaussian approximation evaluating 10 splitpoints

– Gaussian approximation evaluating 100 splitpoints
– Greenwald-Khanna quantile summary with 10 tuples
– Greenwald-Khanna quantile summary with 100 tuples
– Greenwald-Khanna quantile summary with 1000 tuples
– VFML method with 10 bins
– VFML method with 100 bins

31
CHAPTER 7. CLASSIFIERS

– VFML method with 1000 bins

– Exhaustive binary tree

• -e : How many instances between memory consumption checks

• -g : The number of instances a leaf should observe between split

attempts

• -s : Split criterion to use. Example : InfoGainSplitCriterion

• -c : The allowable error in split decision, values closer to 0 will take

longer to decide

• -t : Threshold below which a split will be forced to break ties

• -b : Only allow binary splits

• -z : Stop growing as soon as memory limit is hit

• -r : Disable poor attributes

• -p : Disable pre-pruning

7.1.5 HoeffdingTreeNB
Decision tree for streaming data with Naive Bayes classification at leaves.

• Same parameters as HoeffdingTree

• -q : The number of instances a leaf should observe before permitting

Naive Bayes

7.1.6 HoeffdingTreeNBAdaptive
Decision tree for streaming data with adaptive Naive Bayes classification
at leaves. This adaptive Naive Bayes prediction method monitors the error
rate of majority class and Naive Bayes decisions in every leaf, and chooses
to employ Naive Bayes decisions only where they have been more accu-
rate in past cases.

• Same parameters as HoeffdingTreeNB

32
 7.1. CLASSIFIERS FOR STATIC STREAMS

7.1.7 HoeffdingOptionTree
Decision option tree for streaming data
Hoeffding Option Trees are regular Hoeffding trees containing additional
option nodes that allow several tests to be applied, leading to multiple
Hoeffding trees as separate paths. They consist of a single structure that
efficiently represents multiple trees. A particular example can travel down
multiple paths of the tree, contributing, in different ways, to different op-
tions.
See for details:

[OP] B. Pfahringer, G. Holmes, and R. Kirkby. New options for hoeffding

trees. In AI, pages 90–99, 2007.

Parameters:

• -o : Maximum number of option paths per node

• -m : Maximum memory consumed by the tree

• -n : Numeric estimator to use :

– Gaussian approximation evaluating 10 splitpoints

– Gaussian approximation evaluating 100 splitpoints
– Greenwald-Khanna quantile summary with 10 tuples
– Greenwald-Khanna quantile summary with 100 tuples
– Greenwald-Khanna quantile summary with 1000 tuples
– VFML method with 10 bins
– VFML method with 100 bins
– VFML method with 1000 bins
– Exhaustive binary tree

• -e : How many instances between memory consumption checks

• -g : The number of instances a leaf should observe between split

attempts

• -s : Split criterion to use. Example : InfoGainSplitCriterion

• -c : The allowable error in split decision, values closer to 0 will take

longer to decide

33
CHAPTER 7. CLASSIFIERS

• -w : The allowable error in secondary split decisions, values closer

to 0 will take longer to decide
• -t : Threshold below which a split will be forced to break ties
• -b : Only allow binary splits
• -z : Memory strategy to use
• -r : Disable poor attributes
• -p : Disable pre-pruning
• -d : File to append option table to.

7.1.8 HoeffdingOptionTreeNB
Decision option tree for streaming data with Naive Bayes classification at
leaves.
Parameters:
• Same parameters as HoeffdingOptionTree
• -q : The number of instances a leaf should observe before permitting
Naive Bayes

7.1.9 HoeffdingTreeOptionNBAdaptive
Decision option tree for streaming data with adaptive Naive Bayes classi-
fication at leaves. This adaptive Naive Bayes prediction method monitors
the error rate of majority class and Naive Bayes decisions in every leaf,
and chooses to employ Naive Bayes decisions only where they have been
more accurate in past cases.
Parameters:
• Same parameters as HoeffdingOptionTreeNB

7.1.10 OzaBag
Incremental on-line bagging of Oza and Russell.
Oza and Russell developed online versions of bagging and boosting
for Data Streams. They show how the process of sampling bootstrap repli-
cates from training data can be simulated in a data stream context. They
observe that the probability that any individual example will be chosen
for a replicate tends to a Poisson(1) distribution.

34
 7.1. CLASSIFIERS FOR STATIC STREAMS

[OR] N. Oza and S. Russell. Online bagging and boosting. In Artificial

Intelligence and Statistics 2001, pages 105–112. Morgan Kaufmann,
2001.

Parameters:
• -l : Classifier to train
• -s : The number of models in the bag

7.1.11 OzaBoost
Incremental on-line boosting of Oza and Russell.
See details in:

[OR] N. Oza and S. Russell. Online bagging and boosting. In Artificial

Intelligence and Statistics 2001, pages 105–112. Morgan Kaufmann,
2001.

For the boosting method, Oza and Russell note that the weighting pro-
cedure of AdaBoost actually divides the total example weight into two
halves – half of the weight is assigned to the correctly classified examples,
and the other half goes to the misclassified examples. They use the Poisson
distribution for deciding the random probability that an example is used
for training, only this time the parameter changes according to the boost-
ing weight of the example as it is passed through each model in sequence.
Parameters:
• -l : Classifier to train
• -s : The number of models to boost
• -p : Boost with weights only; no poisson

7.1.12 OCBoost
Online Coordinate Boosting.
Pelossof et al. presented Online Coordinate Boosting, a new online
boosting algorithm for adapting the weights of a boosted classifier, which
yields a closer approximation to Freund and Schapire’s AdaBoost algo-
rithm. The weight update procedure is derived by minimizing AdaBoost’s
loss when viewed in an incremental form. This boosting method may be
reduced to a form similar to Oza and Russell’s algorithm.
See details in:

35
CHAPTER 7. CLASSIFIERS

[PJ] Raphael Pelossof, Michael Jones, Ilia Vovsha, and Cynthia Rudin.
Online coordinate boosting. 2008.

Example:
OCBoost -l HoeffdingTreeNBAdaptive -e 0.5

Parameters:
• -l : Classifier to train

• -s : The number of models to boost

• -e : Smoothing parameter

7.2 Classifiers for evolving streams

7.2.1 OzaBagASHT
Bagging using trees of different size. The Adaptive-Size Hoeffding Tree
(ASHT) is derived from the Hoeffding Tree algorithm with the following
differences:

• it has a maximum number of split nodes, or size

• after one node splits, if the number of split nodes of the ASHT tree
is higher than the maximum value, then it deletes some nodes to
reduce its size

The intuition behind this method is as follows: smaller trees adapt

more quickly to changes, and larger trees do better during periods with
no or little change, simply because they were built on more data. Trees
limited to size s will be reset about twice as often as trees with a size limit
of 2s. This creates a set of different reset-speeds for an ensemble of such
trees, and therefore a subset of trees that are a good approximation for the
current rate of change. It is important to note that resets will happen all
the time, even for stationary datasets, but this behaviour should not have
a negative impact on the ensemble’s predictive performance.
When the tree size exceeds the maximun size value, there are two dif-
ferent delete options:
• delete the oldest node, the root, and all of its children except the one
where the split has been made. After that, the root of the child not
deleted becomes the new root

36
 7.2. CLASSIFIERS FOR EVOLVING STREAMS

• delete all the nodes of the tree, i.e., restart from a new root.

The maximum allowed size for the n-th ASHT tree is twice the maxi-
mum allowed size for the (n − 1)-th tree. Moreover, each tree has a weight
proportional to the inverse of the square of its error, and it monitors its er-
ror with an exponential weighted moving average (EWMA) with α = .01.
The size of the first tree is 2.
With this new method, it is attempted to improve bagging performance
by increasing tree diversity. It has been observed that boosting tends to
produce a more diverse set of classifiers than bagging, and this has been
cited as a factor in increased performance.
See more details in:

[BHPKG] Albert Bifet, Geoff Holmes, Bernhard Pfahringer, Richard Kirkby,

and Ricard Gavaldà. New ensemble methods for evolving data
streams. In 15th ACM SIGKDD International Conference on Knowledge
Discovery and Data Mining, 2009.

The learner must be ASHoeffdingTree, a Hoeffding Tree with a maxi-

mum size value.
Example:
OzaBagASHT -l ASHoeffdingTree -s 10 -u -r

Parameters:

• Same parameters as OzaBag

• -f : the size of first classifier in the bag.

• -u : Enable weight classifiers

• -r : Reset trees when size is higher than the max

7.2.2 OzaBagADWIN
Bagging using ADWIN. ADWIN is a change detector and estimator that solves
in a well-specified way the problem of tracking the average of a stream
of bits or real-valued numbers. ADWIN keeps a variable-length window
of recently seen items, with the property that the window has the maxi-
mal length statistically consistent with the hypothesis “there has been no
change in the average value inside the window”.
More precisely, an older fragment of the window is dropped if and
only if there is enough evidence that its average value differs from that

37
 CHAPTER 7. CLASSIFIERS

of the rest of the window. This has two consequences: one, that change
reliably declared whenever the window shrinks; and two, that at any time
the average over the existing window can be reliably taken as an estima-
tion of the current average in the stream (barring a very small or very re-
cent change that is still not statistically visible). A formal and quantitative
statement of these two points (a theorem) appears in

[BG07c] Albert Bifet and Ricard Gavaldà. Learning from time-changing data
with adaptive windowing. In SIAM International Conference on Data
Mining, 2007.

ADWIN is parameter- and assumption-free in the sense that it automat-

ically detects and adapts to the current rate of change. Its only parameter
is a confidence bound δ, indicating how confident we want to be in the
algorithm’s output, inherent to all algorithms dealing with random pro-
cesses.
Also important, ADWIN does not maintain the window explicitly, but
compresses it using a variant of the exponential histogram technique. This
means that it keeps a window of length W using only O(log W) memory
and O(log W) processing time per item.
ADWIN Bagging is the online bagging method of Oza and Rusell with
the addition of the ADWIN algorithm as a change detector and as an esti-
mator for the weights of the boosting method. When a change is detected,
the worst classifier of the ensemble of classifiers is removed and a new
classifier is added to the ensemble.
See details in:

[BHPKG] Albert Bifet, Geoff Holmes, Bernhard Pfahringer, Richard Kirkby,

and Ricard Gavaldà. New ensemble methods for evolving data
streams. In 15th ACM SIGKDD International Conference on Knowledge
Discovery and Data Mining, 2009.

Example:

OzaBagAdwin -l HoeffdingTreeNBAdaptive -s 10

Parameters:

• -l : Classifier to train

• -s : The number of models in the bag

38
 7.2. CLASSIFIERS FOR EVOLVING STREAMS

7.2.3 SingleClassifierDrift
Class for handling concept drift datasets with a wrapper on a classifier.
The drift detection method (DDM) proposed by Gama et al. controls
the number of errors produced by the learning model during prediction.
It compares the statistics of two windows: the first one contains all the
data, and the second one contains only the data from the beginning until
the number of errors increases. Their method doesn’t store these windows
in memory. It keeps only statistics and a window of recent errors.
They consider that the number of errors in a sample of examples is
modeled by a binomial distribution. A significant increase in the error of
the algorithm, suggests that the class distribution is changing and, hence,
the actual decision model is supposed to be inappropriate. They check for
a warning level and a drift level. Beyond these levels, change of context is
considered.
The number of errors in a sample of n examples is modeled by a bino-
mial distribution. For each point i in the sequence that is being sampled,
the error rate is thep
probability of misclassifying (pi ), with standard devia-
tion given by si = pi (1 − pi )/i. A significant increase in the error of the
algorithm, suggests that the class distribution is changing and, hence, the
actual decision model is supposed to be inappropriate. Thus, they store
the values of pi and si when pi + si reaches its minimum value during
the process (obtaining ppmin and smin ). And it checks when the following
conditions trigger:

• pi + si ≥ pmin + 2 · smin for the warning level. Beyond this level, the
examples are stored in anticipation of a possible change of context.

• pi + si ≥ pmin + 3 · smin for the drift level. Beyond this level, the
model induced by the learning method is reset and a new model is
learnt using the examples stored since the warning level triggered.

Baena-Garcı́a et al. proposed a new method EDDM in order to im-

prove DDM. It is based on the estimated distribution of the distances be-
tween classification errors. The window resize procedure is governed by
the same heuristics.
See more details in:

[GMCR] J. Gama, P. Medas, G. Castillo, and P. Rodrigues. Learning with drift

detection. In SBIA Brazilian Symposium on Artificial Intelligence, pages
286–295, 2004.

39
 CHAPTER 7. CLASSIFIERS

[BDF] Manuel Baena-Garcı́a, José del Campo-Ávila, Raúl Fidalgo, Albert

Bifet, Ricard Gavaldá, and Rafael Morales-Bueno. Early drift detec-
tion method. In Fourth International Workshop on Knowledge Discovery
from Data Streams, 2006.

Example:
SingleClassifierDrift -d EDDM -l HoeffdingTreeNBAdaptive

Parameters:

• -l : Classifier to train

• -d : Drift detection method to use: DDM or EDDM

7.2.4 AdaHoeffdingOptionTree
Adaptive decision option tree for streaming data with adaptive Naive
Bayes classification at leaves.
An Adaptive Hoeffding Option Tree is a Hoeffding Option Tree with the
following improvement: each leaf stores an estimation of the current error.
It uses an EWMA estimator with α = .2. The weight of each node in the
voting process is proportional to the square of the inverse of the error.
Example:
AdaHoeffdingOptionTree -o 50

Parameters:

• Same parameters as HoeffdingOptionTreeNB

40
 Writing a classifier
8
8.1 Creating a new classifier
To demonstrate the implementation and operation of learning algorithms
in the system, the Java code of a simple decision stump classifier is studied.
The classifier monitors the result of splitting on each attribute and chooses
the attribute the seems to best separate the classes, based on information
gain. The decision is revisited many times, so the stump has potential to
change over time as more examples are seen. In practice it is unlikely to
change after sufficient training.
To describe the implementation, relevant code fragments are discussed
in turn, with the entire code listed (Listing 8.7) at the end. The line num-
bers from the fragments match up with the final listing.
A simple approach to writing a classifier is to extend
moa.classifiers.AbstractClassifier (line 10), which will take
care of certain details to ease the task.

Listing 8.1: Option handling

14 public I n t O p t i o n gracePeriodOption = new I n t O p t i o n ( ” g r a c e P e r i o d ” , ’ g ’ ,
15 ”The number o f i n s t a n c e s t o observe between model changes . ” ,
16 1 0 0 0 , 0 , I n t e g e r .MAX VALUE ) ;
17
18 public FlagOption b i n a r y S p l i t s O p t i o n = new FlagOption ( ” b i n a r y S p l i t s ” , ’ b ’ ,
19 ”Only allow b i n a r y s p l i t s . ” ) ;
20
21 public ClassOption s p l i t C r i t e r i o n O p t i o n = new ClassOption ( ” s p l i t C r i t e r i o n ” ,
22 ’ c ’ , ” S p l i t c r i t e r i o n t o use . ” , S p l i t C r i t e r i o n . c l a s s ,
23 ” InfoGainSplitCriterion ” ) ;

To set up the public interface to the classifier, the options available to

the user must be specified. For the system to automatically take care of
option handling, the options need to be public members of the class, that
extend the moa.options.Option type.
The decision stump classifier example has three options, each of a dif-
ferent type. The meaning of the first three parameters used to construct
options are consistent between different option types. The first parameter

41
 CHAPTER 8. WRITING A CLASSIFIER

is a short name used to identify the option. The second is a character in-
tended to be used on the command line. It should be unique—a command
line character cannot be repeated for different options otherwise an excep-
tion will be thrown. The third standard parameter is a string describing
the purpose of the option. Additional parameters to option constructors
allow things such as default values and valid ranges to be specified.
The first option specified for the decision stump classifier is the “grace
period”. The option is expressed with an integer, so the option has the type
IntOption. The parameter will control how frequently the best stump is
reconsidered when learning from a stream of examples. This increases
the efficiency of the classifier—evaluating after every single example is ex-
pensive, and it is unlikely that a single example will change the decision of
the current best stump. The default value of 1000 means that the choice of
stump will be re-evaluated only after 1000 examples have been observed
since the last evaluation. The last two parameters specify the range of val-
ues that are allowed for the option—it makes no sense to have a negative
grace period, so the range is restricted to integers 0 or greater.
The second option is a flag, or a binary switch, represented by a
FlagOption. By default all flags are turned off, and will be turned on
only when a user requests so. This flag controls whether the decision
stumps should only be allowed to split two ways. By default the stumps
are allowed have more than two branches.
The third option determines the split criterion that is used to decide
which stumps are the best. This is a ClassOption that requires a particu-
lar Java class of the type SplitCriterion. If the required class happens
to be an OptionHandler then those options will be used to configure the
object that is passed in.

Listing 8.2: Miscellaneous fields

25 protected AttributeSplitSuggestion b e s t S p l i t ;
26
27 p r o t e c t e d DoubleVector o b s e r v e d C l a s s D i s t r i b u t i o n ;
28
29 p r o t e c t e d AutoExpandVector<A t t r i b u t e C l a s s O b s e r v e r> a t t r i b u t e O b s e r v e r s ;
30
31 p r o t e c t e d double w e i g h t S e e n A t L a s t S p l i t ;
32
33 public boolean isRandomizable ( ) {
34 return false ;
35 }

Four global variables are used to maintain the state of the classifier.
The bestSplit field maintains the current stump that has been cho-
sen by the classifier. It is of type AttributeSplitSuggestion, a class
used to split instances into different subsets.
The observedClassDistribution field remembers the overall dis-
tribution of class labels that have been observed by the classifier. It is of

42
 8.1. CREATING A NEW CLASSIFIER

type DoubleVector, which is a handy class for maintaining a vector of

floating point values without having to manage its size.
The attributeObservers field stores a collection of
AttributeClassObservers, one for each attribute. This is the informa-
tion needed to decide which attribute is best to base the stump on.
The weightSeenAtLastSplit field records the last time an evalua-
tion was performed, so that it can be determined when another evaluation
is due, depending on the grace period parameter.
The isRandomizable() function needs to be implemented to specify
whether the classifier has an element of randomness. If it does, it will
automatically be set up to accept a random seed. This classifier is does
not, so false is returned.

Listing 8.3: Preparing for learning

37 @Override
38 public void r e s e t L e a r n i n g I m p l ( ) {
39 t h i s . b e s t S p l i t = null ;
40 t h i s . o b s e r v e d C l a s s D i s t r i b u t i o n = new DoubleVector ( ) ;
41 t h i s . a t t r i b u t e O b s e r v e r s = new AutoExpandVector<A t t r i b u t e C l a s s O b s e r v e r > ( ) ;
42 this . weightSeenAtLastSplit = 0 . 0 ;
43 }

This function is called before any learning begins, so it should set the
default state when no information has been supplied, and no training ex-
amples have been seen. In this case, the four global fields are set to sensible
defaults.

Listing 8.4: Training on examples

45 @Override
46 public void t r a i n O n I n s t a n c e I m p l ( I n s t a n c e i n s t ) {
47 t h i s . o b s e r v e d C l a s s D i s t r i b u t i o n . addToValue ( ( i n t ) i n s t . c l a s s V a l u e ( ) , i n s t
48 . weight ( ) ) ;
49 f o r ( i n t i = 0 ; i < i n s t . numAttributes ( ) − 1 ; i ++) {
50 i n t i n s t A t t I n d e x = modelAttIndexToInstanceAttIndex ( i , i n s t ) ;
51 A t t r i b u t e C l a s s O b s e r v e r obs = t h i s . a t t r i b u t e O b s e r v e r s . g e t ( i ) ;
52 i f ( obs == n u l l ) {
53 obs = i n s t . a t t r i b u t e ( i n s t A t t I n d e x ) . isNominal ( ) ?
54 newNominalClassObserver ( ) : newNumericClassObserver ( ) ;
55 t h i s . a t t r i b u t e O b s e r v e r s . s e t ( i , obs ) ;
56 }
57 obs . o b s e r v e A t t r i b u t e C l a s s ( i n s t . value ( i n s t A t t I n d e x ) , ( i n t ) i n s t
58 . c l a s s V a l u e ( ) , i n s t . weight ( ) ) ;
59 }
60 i f ( t h i s . trainingWeightSeenByModel − t h i s . w e i g h t S e e n A t L a s t S p l i t >=
61 t h i s . gracePeriodOption . getValue ( ) ) {
62 this . bestSplit = findBestSplit ( ( SplitCriterion )
63 getPreparedClassOption ( t h i s . s p l i t C r i t e r i o n O p t i o n ) ) ;
64 t h i s . w e i g h t S e e n A t L a s t S p l i t = t h i s . trainingWeightSeenByModel ;
65 }
66 }

This is the main function of the learning algorithm, called for every
training example in a stream. The first step, lines 47-48, updates the over-
all recorded distribution of classes. The loop from line 49 to line 59 repeats
for every attribute in the data. If no observations for a particular attribute
have been seen previously, then lines 53-55 create a new observing object.
Lines 57-58 update the observations with the values from the new exam-

43
 CHAPTER 8. WRITING A CLASSIFIER

ple. Lines 60-61 check to see if the grace period has expired. If so, the best
split is re-evaluated.
Listing 8.5: Functions used during training
79 p r o t e c t e d A t t r i b u t e C l a s s O b s e r v e r newNominalClassObserver ( ) {
80 r e t u r n new NominalAttributeClassObserver ( ) ;
81 }
82
83 p r o t e c t e d A t t r i b u t e C l a s s O b s e r v e r newNumericClassObserver ( ) {
84 r e t u r n new GaussianNumericAttributeClassObserver ( ) ;
85 }
86
87 protected AttributeSplitSuggestion f i n d B e s t S p l i t ( S p l i t C r i t e r i o n c r i t e r i o n ) {
88 A t t r i b u t e S p l i t S u g g e s t i o n bestFound = n u l l ;
89 double b e s t M e r i t = Double . NEGATIVE INFINITY ;
90 double [ ] p r e S p l i t D i s t = t h i s . o b s e r v e d C l a s s D i s t r i b u t i o n . getArrayCopy ( ) ;
91 f o r ( i n t i = 0 ; i < t h i s . a t t r i b u t e O b s e r v e r s . s i z e ( ) ; i ++) {
92 A t t r i b u t e C l a s s O b s e r v e r obs = t h i s . a t t r i b u t e O b s e r v e r s . g e t ( i ) ;
93 i f ( obs ! = n u l l ) {
94 AttributeSplitSuggestion suggestion =
95 obs . g e t B e s t E v a l u a t e d S p l i t S u g g e s t i o n (
96 criterion ,
97 preSplitDist ,
98 i,
99 this . binarySplitsOption . i s S e t ( ) ) ;
100 i f ( suggestion . merit > bestMerit ) {
101 bestMerit = suggestion . merit ;
102 bestFound = s u g g e s t i o n ;
103 }
104 }
105 }
106 r e t u r n bestFound ;
107 }

These functions assist the training algorithm.

newNominalClassObserver and newNumericClassObserver are re-
sponsible for creating new observer objects for nominal and numeric at-
tributes, respectively. The findBestSplit() function will iterate through
the possible stumps and return the one with the highest ‘merit’ score.
Listing 8.6: Predicting class of unknown examples
68 public double [ ] g e t V o t e s F o r I n s t a n c e ( I n s t a n c e i n s t ) {
69 i f ( t h i s . b e s t S p l i t != null ) {
70 i n t branch = t h i s . b e s t S p l i t . s p l i t T e s t . b r a n c h F o r I n s t a n c e ( i n s t ) ;
71 i f ( branch >= 0 ) {
72 return this . b e s t S p l i t
73 . r e s u l t i n g C l a s s D i s t r i b u t i o n F r o m S p l i t ( branch ) ;
74 }
75 }
76 r e t u r n t h i s . o b s e r v e d C l a s s D i s t r i b u t i o n . getArrayCopy ( ) ;
77 }

This is the other important function of the classifier besides training—

using the model that has been induced to predict the class of examples. For
the decision stump, this involves calling the functions branchForInstance()
and resultingClassDistributionFromSplit() that are implemented
by the AttributeSplitSuggestion class.
Putting all of the elements together, the full listing of the tutorial class
is given below.
Listing 8.7: Full listing
1 package moa . c l a s s i f i e r s ;
2
3 import moa . c o r e . AutoExpandVector ;

44
 8.1. CREATING A NEW CLASSIFIER

4 import moa . c o r e . DoubleVector ;

5 import moa . o p t i o n s . ClassOption ;
6 import moa . o p t i o n s . FlagOption ;
7 import moa . o p t i o n s . I n t O p t i o n ;
8 import weka . c o r e . I n s t a n c e ;
9
10 public c l a s s D e c i s i o n S t u m p T u t o r i a l extends A b s t r a c t C l a s s i f i e r {
11
12 p r i v a t e s t a t i c f i n a l long s e r i a l V e r s i o n U I D = 1L ;
13
14 public I n t O p t i o n gracePeriodOption = new I n t O p t i o n ( ” g r a c e P e r i o d ” , ’ g ’ ,
15 ”The number o f i n s t a n c e s t o observe between model changes . ” ,
16 1 0 0 0 , 0 , I n t e g e r .MAX VALUE ) ;
17
18 public FlagOption b i n a r y S p l i t s O p t i o n = new FlagOption ( ” b i n a r y S p l i t s ” , ’ b ’ ,
19 ”Only allow b i n a r y s p l i t s . ” ) ;
20
21 public ClassOption s p l i t C r i t e r i o n O p t i o n = new ClassOption ( ” s p l i t C r i t e r i o n ” ,
22 ’ c ’ , ” S p l i t c r i t e r i o n t o use . ” , S p l i t C r i t e r i o n . c l a s s ,
23 ” InfoGainSplitCriterion ” ) ;
24
25 protected AttributeSplitSuggestion b e s t S p l i t ;
26
27 p r o t e c t e d DoubleVector o b s e r v e d C l a s s D i s t r i b u t i o n ;
28
29 p r o t e c t e d AutoExpandVector<A t t r i b u t e C l a s s O b s e r v e r> a t t r i b u t e O b s e r v e r s ;
30
31 p r o t e c t e d double w e i g h t S e e n A t L a s t S p l i t ;
32
33 public boolean isRandomizable ( ) {
34 return false ;
35 }
36
37 @Override
38 public void r e s e t L e a r n i n g I m p l ( ) {
39 t h i s . b e s t S p l i t = null ;
40 t h i s . o b s e r v e d C l a s s D i s t r i b u t i o n = new DoubleVector ( ) ;
41 t h i s . a t t r i b u t e O b s e r v e r s = new AutoExpandVector<A t t r i b u t e C l a s s O b s e r v e r > ( ) ;
42 this . weightSeenAtLastSplit = 0 . 0 ;
43 }
44
45 @Override
46 public void t r a i n O n I n s t a n c e I m p l ( I n s t a n c e i n s t ) {
47 t h i s . o b s e r v e d C l a s s D i s t r i b u t i o n . addToValue ( ( i n t ) i n s t . c l a s s V a l u e ( ) , i n s t
48 . weight ( ) ) ;
49 f o r ( i n t i = 0 ; i < i n s t . numAttributes ( ) − 1 ; i ++) {
50 i n t i n s t A t t I n d e x = modelAttIndexToInstanceAttIndex ( i , i n s t ) ;
51 A t t r i b u t e C l a s s O b s e r v e r obs = t h i s . a t t r i b u t e O b s e r v e r s . g e t ( i ) ;
52 i f ( obs == n u l l ) {
53 obs = i n s t . a t t r i b u t e ( i n s t A t t I n d e x ) . isNominal ( ) ?
54 newNominalClassObserver ( ) : newNumericClassObserver ( ) ;
55 t h i s . a t t r i b u t e O b s e r v e r s . s e t ( i , obs ) ;
56 }
57 obs . o b s e r v e A t t r i b u t e C l a s s ( i n s t . value ( i n s t A t t I n d e x ) , ( i n t ) i n s t
58 . c l a s s V a l u e ( ) , i n s t . weight ( ) ) ;
59 }
60 i f ( t h i s . trainingWeightSeenByModel − t h i s . w e i g h t S e e n A t L a s t S p l i t >=
61 t h i s . gracePeriodOption . getValue ( ) ) {
62 this . bestSplit = findBestSplit ( ( SplitCriterion )
63 getPreparedClassOption ( t h i s . s p l i t C r i t e r i o n O p t i o n ) ) ;
64 t h i s . w e i g h t S e e n A t L a s t S p l i t = t h i s . trainingWeightSeenByModel ;
65 }
66 }
67
68 public double [ ] g e t V o t e s F o r I n s t a n c e ( I n s t a n c e i n s t ) {
69 i f ( t h i s . b e s t S p l i t != null ) {
70 i n t branch = t h i s . b e s t S p l i t . s p l i t T e s t . b r a n c h F o r I n s t a n c e ( i n s t ) ;
71 i f ( branch >= 0 ) {
72 return this . b e s t S p l i t
73 . r e s u l t i n g C l a s s D i s t r i b u t i o n F r o m S p l i t ( branch ) ;
74 }
75 }
76 r e t u r n t h i s . o b s e r v e d C l a s s D i s t r i b u t i o n . getArrayCopy ( ) ;
77 }
78
79 p r o t e c t e d A t t r i b u t e C l a s s O b s e r v e r newNominalClassObserver ( ) {
80 r e t u r n new NominalAttributeClassObserver ( ) ;
81 }
82
83 p r o t e c t e d A t t r i b u t e C l a s s O b s e r v e r newNumericClassObserver ( ) {
84 r e t u r n new GaussianNumericAttributeClassObserver ( ) ;
85 }
86

45
 CHAPTER 8. WRITING A CLASSIFIER

87 protected AttributeSplitSuggestion f i n d B e s t S p l i t ( S p l i t C r i t e r i o n c r i t e r i o n ) {
88 A t t r i b u t e S p l i t S u g g e s t i o n bestFound = n u l l ;
89 double b e s t M e r i t = Double . NEGATIVE INFINITY ;
90 double [ ] p r e S p l i t D i s t = t h i s . o b s e r v e d C l a s s D i s t r i b u t i o n . getArrayCopy ( ) ;
91 f o r ( i n t i = 0 ; i < t h i s . a t t r i b u t e O b s e r v e r s . s i z e ( ) ; i ++) {
92 A t t r i b u t e C l a s s O b s e r v e r obs = t h i s . a t t r i b u t e O b s e r v e r s . g e t ( i ) ;
93 i f ( obs ! = n u l l ) {
94 AttributeSplitSuggestion suggestion =
95 obs . g e t B e s t E v a l u a t e d S p l i t S u g g e s t i o n (
96 criterion ,
97 preSplitDist ,
98 i,
99 this . binarySplitsOption . i s S e t ( ) ) ;
100 i f ( suggestion . merit > bestMerit ) {
101 bestMerit = suggestion . merit ;
102 bestFound = s u g g e s t i o n ;
103 }
104 }
105 }
106 r e t u r n bestFound ;
107 }
108
109 public void g et M od e lD e sc r ip t io n ( S t r i n g B u i l d e r out , i n t i n d e n t ) {
110 }
111
112 p r o t e c t e d moa . c o r e . Measurement [ ] getModelMeasurementsImpl ( ) {
113 return null ;
114 }
115
116 }

8.2 Compiling a classifier

The following five files are assumed to be in the current working directory:

DecisionStumpTutorial.java
moa.jar
weka.jar
sizeofag.jar

The example source code can be compiled with the following com-
mand:

javac -cp moa.jar:weka.jar DecisionStumpTutorial.java

This produces compiled java class file DecisionStumpTutorial.class.

Before continuing, the commands below set up directory structure to
reflect the package structure:

mkdir moa
mkdir moa/classifiers
cp DecisionStumpTutorial.class moa/classifiers/

The class is now ready to use.

46
Bi-directional interface with WEKA
9
Now, it is easy to use MOA classifiers and streams from WEKA, and WEKA
classifiers from MOA. The main difference between using incremental clas-
sifiers in WEKA and in MOA will be the evaluation method used.

9.1 WEKA classifiers from MOA

Weka classifiers may be incremental or non incremental methods.

To use the Weka classifiers from MOA it is necessary to use one of the
following classes:

47
CHAPTER 9. BI-DIRECTIONAL INTERFACE WITH WEKA

9.1.1 WekaClassifier
A classifier to use classifiers from WEKA.

Wi W W

T T
WEKAClassifier builds a model of W instances every T instances only
for non incremental methods. For WEKA incremental methods the WEKA
classifier is trained for every instance.
Example:
WekaClassifier -l weka.classifiers.trees.J48
-w 10000 -i 1000 -f 100000

Parameters:

• -l : Classifier to train

• -w : Size of Window for training learner

• -i : Size of first Window for training learner

• -f : How many instances between samples of the learning perfor-

mance

9.1.2 SingleClassifierDrift
Class for handling concept drift datasets with a wrapper on a classifier.
The drift detection method (DDM) proposed by Gama et al. controls
the number of errors produced by the learning model during prediction.
It compares the statistics of two windows: the first one contains all the
data, and the second one contains only the data from the beginning until
the number of errors increases. Their method doesn’t store these windows
in memory. It keeps only statistics and a window of recent errors.
They consider that the number of errors in a sample of examples is
modeled by a binomial distribution. A significant increase in the error of
the algorithm, suggests that the class distribution is changing and, hence,
the actual decision model is supposed to be inappropriate. They check for
a warning level and a drift level. Beyond these levels, change of context is
considered.

48
 9.1. WEKA CLASSIFIERS FROM MOA

The number of errors in a sample of n examples is modeled by a bino-

mial distribution. For each point i in the sequence that is being sampled,
the error rate is thep
probability of misclassifying (pi ), with standard devia-
tion given by si = pi (1 − pi )/i. A significant increase in the error of the
algorithm, suggests that the class distribution is changing and, hence, the
actual decision model is supposed to be inappropriate. Thus, they store
the values of pi and si when pi + si reaches its minimum value during
the process (obtaining ppmin and smin ). And it checks when the following
conditions trigger:

• pi + si ≥ pmin + 2 · smin for the warning level. Beyond this level, the
examples are stored in anticipation of a possible change of context.

• pi + si ≥ pmin + 3 · smin for the drift level. Beyond this level, the
model induced by the learning method is reset and a new model is
learnt using the examples stored since the warning level triggered.

Baena-Garcı́a et al. proposed a new method EDDM in order to im-

prove DDM. It is based on the estimated distribution of the distances be-
tween classification errors. The window resize procedure is governed by
the same heuristics.
See more details in:

[GMCR] J. Gama, P. Medas, G. Castillo, and P. Rodrigues. Learning with drift

detection. In SBIA Brazilian Symposium on Artificial Intelligence, pages
286–295, 2004.

[BDF] Manuel Baena-Garcı́a, José del Campo-Ávila, Raúl Fidalgo, Albert

Bifet, Ricard Gavaldá, and Rafael Morales-Bueno. Early drift detec-
tion method. In Fourth International Workshop on Knowledge Discovery
from Data Streams, 2006.

Example:

SingleClassifierDrift -d EDDM
-l weka.classifiers.bayes.NaiveBayesUpdateable

Parameters:

• -l : Classifier to train

• -d : Drift detection method to use: DDM or EDDM

49
CHAPTER 9. BI-DIRECTIONAL INTERFACE WITH WEKA

9.2 MOA classifiers from WEKA

You can use MOA classifiers quite easily as incremental classifiers within
the WEKA Explorer, Knowledge Flow interface or command-line inter-
face, using the weka.classifiers.meta.MOA meta-classifier. This meta-
classifier is just a wrapper for MOA classifiers, translating the WEKA method
calls into MOA ones.

You can use MOA streams within the WEKA framework using the
weka.datagenerators.classifiers.classification.MOA data gen-
erator. For example:
weka.datagenerators.classifiers.classification.MOA
-B moa.streams.generators.LEDGenerator

In order to manipulate the MOA classifiers in the GUI, like Explorer or

Experimenter, you need to register a new editor for the GenericObjectEdi-
tor.
The following steps are necessary to integrate the MOA classifiers:
1. Determine the location of your home directory:
• Windows in a command prompt, run the following command:
echo %USERPROFILE%
• Linux/Unix in a terminal (bash), run the following command:
echo $HOME
2. Copy the ”GUIEdtitors.props.addon” file, contained in the MOA project
in the source directory ”src/main/java/weka/gui” or in the moa.jar
in ”weka/gui”, into your home directory. Rename the extension
from ”.props.addon” to ”.props”. If the home directory already con-
tains such a file, then just append the content of the version in MOA
file to the already existing one.

50
 9.2. MOA CLASSIFIERS FROM WEKA

3. Restart WEKA from the MOA project, e.g., using the ”run-explorer”
target of the ANT build file.

51