Scribd
Upload
30 views

Problem - Decision Tree Implementation

This document provides instructions for a decision tree implementation project using the Iris dataset. The project has two parts: 1. Print the steps of building a decision tree, including entropy calculations and splits at each level. Sample expected output is provided. 2. Build an actual decision tree classifier and print the tree. Extra marks are awarded for this part. The submission should include a Jupyter notebook with Python code, comments, and any additional images. Scoring will be based on correctness, code quality, commenting, and a bonus for implementing the full decision tree model.

Uploaded by

Vijayray
Copyright
© © All Rights Reserved
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
30 views

Problem - Decision Tree Implementation

This document provides instructions for a decision tree implementation project using the Iris dataset. The project has two parts: 1. Print the steps of building a decision tree, including entropy calculations and splits at each level. Sample expected output is provided. 2. Build an actual decision tree classifier and print the tree. Extra marks are awarded for this part. The submission should include a Jupyter notebook with Python code, comments, and any additional images. Scoring will be based on correctness, code quality, commenting, and a bonus for implementing the full decision tree model.

Uploaded by

Vijayray
Copyright
© © All Rights Reserved
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 2

Decision Tree Implementation

For Iris dataset your task is to print steps for every split in the decision tree.
Your Project is divided in 2 parts -
1. Print the Decision tree steps as specified in the example below. This is of 80 marks in total.

2. Decision Tree Implementation: Building actual decision tree for the dataset and then printing
it. This is of 20 marks.

Note: You can submit multiple times before deadline/evaluation. Your final submission
will be evaluated. No submission is allowed after the project is evaluated by TA.
Dataset: Iris dataset.
Implementation: Consider the decision tree for OR below:

Expected Output:
Level 0

Count of 0(False) = 1

Count of 1(True) = 3

Current Entropy is = 0.811278124459

Splitting on feature X1 with gain ratio 0.311278124459

Level 1
Count of 0 = 1

Count of 1 = 1

Current Entropy is = 1.0

Splitting on feature X2 with gain ratio 1.0

Level 2

Count of 0 = 1

Current Entropy is = 0.0

Reached leaf Node

Level 2

Count of 1 = 1

Current Entropy is = 0.0

Reached leaf Node

Level 1

Count of 1 = 2

Current Entropy is = 0.0

Reached leaf Node

Comments : Your code must have proper comments for better understanding.
Extra features : You can add extra images etc for clarity and better presentation of
your results.
Score : Score will be given by the TA based on your submission.
Submission : You have to upload zipped file which has python notebook with
implementation.
Your project will be evaluated on following parameters -
· Correctness (Max Score 60)
· Quality of Code (Max Score 10)
· Proper Commenting (Max Score 10)
· Bonus: Actual Tree Implementation instead of just printing steps.  (Max Score 20)

You might also like