UNIT 5 : Introduction to Platforms, Tools, Frameworks and Languages for

AIML

1. Top AIML Tools, Software and Programming Languages

1. Salesforce Einstein: The Salesforce Einstein is an AI platform that creates

recommendations and predictions based on business process inputs and
customer data. This platform is used to automate customer-facing responses to
optimise productivity and customer relationship.
2. IBM Watson: IBM Watson is a highly efficient AI software that is capable of
learning more with very few data points. Companies even leverage IBM’s
industry expertise to optimise their own business processes. This software
allows businesses to build models from scratch or use an existing pre-trained
business solution.
3. Deep Vision: This facial recognition model monitors demographic variation
for any geographic location to understand behaviour. Companies use this model
to decode customer patterns and design their marketing campaigns accordingly.
This model can also be used for predictive analysis and target potential leads.
4. Cloud Machine Learning Engine: A cloud-based predictive analytics
modelling platform, Cloud Machine Learning Engine works on both large and
small size data to provide insights on different business processes. Companies
have built secure infrastructures, APIs, data analytics, developer and machine
learning tools using this model.
5. Azure Machine Learning Studio: This provides an interactive platform for
predictive analysis. The drag-and-drop feature provides an easy to use interface
to create innovative cloud-based apps and edge computing. Azure Machine
Learning Studio also offers infrastructure, AI services and tools for developing
intelligent solutions.
6. Nvidia Deep Learning AI: A suite of products dedicated to deep learning and
machine intelligence, this software helps organisations with smart and
predictive analysis. Used by government and businesses alike, the Nvidia deep
learning AI provides an adaptive platform for all kinds of analytical insights.
7. Playment: Playment is a data labelling platform that prepares training data for
computer vision models at scale. It specialises in data relating to autonomous
vehicle, drones, mapping, and similar spaces which require high precision
annotation.

1. Machine Learning tools

1. TensorFlow: TensorFlow has topped the charts for machine learning tools in
2022. It offers a JS library that helps to develop and train machine learning
algorithms. This open-source machine learning library was created by Google.
It is flexible, easily adaptable and has an exhaustive list of tools for machine
learning training models.
2. Amazon Machine Learning: This is a cloud-based machine learning
software which can be used at scale by web or mobile app developers. Amazon
machine learning is widely used for creating machine learning models and
predictive analysis. It provides wizards and visualization tools and supports
three types of models (multi-class classification, binary classification and
regression).
 3. Accord.NET: This .Net ML framework has image and audio processing
libraries in C#, making it a favourite among developers. It includes multiple
libraries for a number of applications like linear algebra, statistical data
processing, pattern recognition and more.
4. Apache Mahout: This is a framework for Scala DSL and distributed linear
algebra. An open-source free project of the Apache software foundation, this
framework can implement algorithms for mathematicians, statisticians and data
scientists.
5. Shogun: This machine learning library has C++ programming languages
which provides data structures for ML problems. Additionally, it supports other
languages like Ruby, Lua, Java, Octave, R, Python and more so that it is widely
adaptable.
2. Programming Languages
Python
Python is one of the top programming languages due to its simplicity. The syntax and
data structure of Python is easy to learn. Object-oriented programming, functional
programming as well as procedure-oriented style of programming by Python. A
number of libraries exist within this programming language, making our tasks simpler.
Being one of them. Helps us settle numerous logical calculations. Python offers rapid
prototyping compared to other programming languages such as Java or C++. It can be
used in a numerous AIML and requires a smaller number of code lines compared to
other programming languages.

R
R is a full-fledged software that is used for statistical computing and visualization, it
is a free and open source programming language, allowing it to be easily accessed.
Programmers can learn from the source code and modify it as required. R has been in
use for a long time now and has been used in areas such as and, It is one of the best
programming languages when it comes to statistical purposes as is useful for
controlling and analyzing data. It has various packages like RODBC, and these are
used in the field of AI and ML. These packages allow the language to implement ML
Algorithms easily. R also has a wide utilization is the banking sector.

Java
The existence of Java dates back way before Python, and it works well for Machine
Learning development. It provides many benefits such as, ease of use, package
services, simplified work with large-scale projects, graphical representation of data
and better user interaction. Game development and mobile application development
can also be done using Java. One of the most popular toolkits for Java is Weka, this
contains tools for data pre-processing, classification, regression and visualization.
Another library called Deeplearning4j which is specially designed for Java, provides
support for deep learning algorithms.

Julia
Compared to the other programming languages, Julia is a newer, high-level, free
programming language. It is rapidly being adopted by the tech as well as finance
community due to its ability to execute ML Algorithms at a fast speed. Investment
banks are using Julia for time series analytics and insurance companies are using it for
risk calculation. One of the major advantages is that Julia is equipped with Flux. Flux
provides a flexible interface which can be modified to create advanced models. Some
of the tools available for Julia are a linter, profiler, debugger package and many more.
Julia is adaptable to existing ML frameworks such as TensorFlow and MXNet.

C/C++
C++ is a general-purpose language developed as an extension for the popular C
Programming Language. It was planned with an inclination towards
resource-constrained software and large systems, effectiveness and adaptability of
utilization as its main features. Most people start with C/C++ as their first
programming language. C/C++ is used in the area of AI in Gaming. It has
sophisticated libraries such as Mlpack, which is a fast and flexible ML Library. The
main aim is to implement ML Algorithms at a faster speed.

List of AI Tools & Frameworks

TensorFlow. It's an open-source AI library for dataflow and differentiable
programming developed by the Google Brain team. ...
Sci-kit Learn. A free machine learning library that integrates seamlessly with
the Python language. ...
Pandas. ...
NumPy. ...
PyTorch. ...
Keras. ...
Theano. ...
Google ML Kit.

2. Machine Learning Tools

TensorFlow

It’s an open-source AI library for dataflow and differentiable programming

developed by the Google Brain team. It's primarily used for machine learning
applications and deep neural network research.

Languages: Python, C++, and Java

Platforms: Linux, macOS, Windows, Android, iOS (via TensorFlow Lite)

+ Highly flexible and modular;

+ Excellent community support;

+ Scalable for both small and large applications;

- The steeper learning curve for beginners;

- Some operations can be less intuitive compared to other frameworks.

“TensorFlow, as one of the market leaders, is ideally suited for large

projects due to a developed ecosystem of products (TensorFlow Hub,
TensorFlow Serving, TensorFlow Lite), great community, detailed
documentation, and lots of code examples over the internet.”

Best softwares for Machine Learning

Google TensorFlow. ...
Amazon Machine Learning. ...
OpenNN. ...
PyTorch. ...
Vertex AI. ...
BigML. ...
Apache Mahout. ...
Weka. Weka is a popular open-source machine learning tool that provides a
collection of algorithms for data preprocessing, classification, regression, clustering,
and visualization.

3. Others

Sci-kit Learn

A free machine learning library that integrates seamlessly with

the Python language. It's suitable for various machine-learning tasks, including
classification, regression, and clustering.

Languages: Python

Platforms: Linux, macOS, Windows

+ Comprehensive library for machine learning algorithms;

+ Strong documentation and community support;

+ Ease of use and integration;

- Not tailored for deep learning;

- Doesn't support GPU acceleration.

“Ski-kit Learn, just like Pandas, is very convenient, and has a bunch of
ready-made packaged solutions.”
Theano

An open-source numerical computation library that lets developers efficiently

define, optimize, and evaluate mathematical expressions.

Languages: Python

Platforms: Linux, macOS X, Windows

+ Highly efficient for numerical tasks;

+ GPU acceleration support;

+ Tight integration with NumPy;

- Development has been halted;

- Not as beginner-friendly as some other options.

Caffe 2

Now we move on to deep learning tools and frameworks. The first one is Caffe
2: an open-source deep learning framework with modularity and speed in mind.
Developed by Facebook, it's the successor to the Caffe framework.

Languages: Python, C++

Platforms: Linux, macOS X, Windows, Android, iOS

+ Highly modular and performant;

+ Suitable for mobile deployment;

+ Strong community and support;

- The steeper learning curve for beginners;

- Less popular compared to TensorFlow and PyTorch.

Keras

It is an open-source neural network library written in Python. It's known for

being user-friendly and modular, acting as an interface for TensorFlow and
Theano.

Languages: Python
Platforms: Linux, macOS X, Windows (depends on the backend: TensorFlow,
Theano, etc.)

+ User-friendly API;

+ Modular and extendable;

+ Supports multiple backend neural computation engines;

- Performance can sometimes be suboptimal;

- Dependency on TensorFlow or Theano as a backend.

“Keras can run on top of TensorFlow or Theano. Keras is, in fact, a

top-level API for these backends (TensorFlow/Theano), which simplifies
model development.”

PyTorch

Developed by Facebook's AI Research lab, it's a dynamic computational

graph-based AI framework often compared to TensorFlow. It allows developers
to use GPU acceleration for computations.

Languages: Python, C++

Platfroms: Linux, macOS X, Windows

+ Dynamic computational graph;

+ Intuitive and more pythonic;

+ Strong support for GPU acceleration;

- Slower than some counterparts due to dynamic computation.

“PyTorch, another market leader with a large community and lots of code
examples, is ideal for smaller projects and for research.”

PyTorch

Developed by Facebook's AI Research lab, it's a dynamic computational

graph-based AI framework often compared to TensorFlow. It allows developers
to use GPU acceleration for computations.

Languages: Python, C++

Platfroms: Linux, macOS X, Windows

+ Dynamic computational graph;

+ Intuitive and more pythonic;

+ Strong support for GPU acceleration;

- Slower than some counterparts due to dynamic computation.

“PyTorch, another market leader with a large community and lots of code
examples, is ideal for smaller projects and for research.”

Microsoft CNTK

The Microsoft Cognitive Toolkit, or CNTK for short, is a deep learning

framework developed by Microsoft. It is known for its efficiency at scale.

Languages: Python, C++

Platforms: Linux, macOS X, Windows

+ Efficient for large-scale datasets;

+ Strong support for recurrent neural networks (RNNs);

+ High performance and scalability;

+ Less intuitive API compared to some counterparts;

+ Limited community support.

MxNet

A deep learning framework designed for both efficiency and flexibility. It

allows developers to mix symbolic and imperative programming.

Languages: Python, C++, Julia, R, Scala, Perl

Platforms: Linux, macOS X, Windows, iOS, Android

+ Supports multi-language APIs;

+ Highly efficient and scalable;

+ Mix of symbolic and imperative programming;

+ Lesser community support compared to TensorFlow and PyTorch;

+ Some parts of the documentation might be lacking.

AI Frameworks Comparision

With a plethora of options available, it's crucial to choose the right one that
aligns with your project requirements, expertise, and platform preferences. In
the table below, we've collated the main aspects of some of the leading AI tools,
providing a side-by-side comparison to aid in your decision-making process.
Let's get a birds-eye view of these tools and discern the subtle nuances that set
each apart.
Google ML Kit

A mobile SDK that brings Google's machine learning capabilities to Android

and iOS apps, with powerful, yet easy-to-use solutions.

Languages: Java (Android), Swift/Objective-C (iOS)

Platforms: Android, iOS

+ Easy to integrate into mobile apps;

+ Pre-trained models available;

+ Supports custom TensorFlow Lite models;

- Limited to mobile app development;

- Some advanced use cases may need further fine-tuning.

Artificial Intelligence Tools & Frameworks

The development of neural networks is a long process that requires a lot of thought
behind the architecture and a whole bunch of nuances that actually make up the
system.

These nuances can easily end up getting overwhelming and not everything can be
easily tracked. Hence, the need for such tools arises, where humans handle the major
architectural decisions leaving other optimization tasks to such tools. Imagine an
architecture with just 4 possible boolean hyperparameters, testing all possible
combinations would take 4! Runs. Retraining the same architecture 24 times is
definitely not the best use of time and energy.

Scikit Learn

Scikit-learn is one of the most well-known ML libraries. It underpins many

administered and unsupervised learning calculations. Precedents incorporate direct
and calculated relapses, choice trees, bunching, k-implies, etc.

It expands on two essential libraries of Python, NumPy and SciPy.

It includes a lot of calculations for regular AI and data mining assignments,

including bunching, relapse and order. Indeed, even undertakings like
changing information, feature determination and ensemble techniques can be
executed in a couple of lines.

For a fledgeling in ML, Scikit-learn is a more-than-adequate instrument to

work with, until you begin actualizing progressively complex calculations.

It utilizes an arrangement of multi-layered hubs that enables you to rapidly set up,
train, and send counterfeit neural systems with huge datasets. This is the thing that
enables Google to recognize questions in photographs or comprehend verbally
expressed words in its voice-acknowledgment application.
Theano

Theano is wonderfully folded over Keras, an abnormal state neural systems library,
that runs nearly in parallel with the Theano library. Keras’ fundamental favorable
position is that it is a moderate Python library for profound discovering that can keep
running over Theano or TensorFlow.

It was created to make actualizing profound learning models as quick and

simple as feasible for innovative work.
It keeps running on Python 2.7 or 3.5 and can consistently execute on GPUs
and CPUs.

What sets Theano separated is that it exploits the PC’s GPU. This enables it to make
information escalated counts up to multiple times quicker than when kept running on
the CPU alone. Theano’s speed makes it particularly profitable for profound learning
and other computationally complex undertakings.

Caffe

‘Caffe’ is a profound learning structure made with articulation, speed, and measured
quality as a top priority. It is created by the Berkeley Vision and Learning Center
(BVLC) and by network donors. Google’s DeepDream depends on Caffe Framework.
This structure is a BSD-authorized C++ library with Python Interface.
MxNet

It allows for trading computation time for memory via ‘forgetful backprop’ which can
be very useful for recurrent nets on very long sequences.

Built with scalability in mind (fairly easy-to-use support for multi-GPU and
multi-machine training).
Lots of cool features, like easily writing custom layers in high-level languages

Unlike almost all other major frameworks, it is not directly governed by a

major corporation which is a healthy situation for an opensource,
community-developed framework.
TVM support, which will further improve deployment support, and allow
running on a whole host of new device types

Keras

If you like the Python-way of doing things, Keras is for you. It is a high-level library
for neural networks, using TensorFlow or Theano as its backend.

The majority of practical problems are more like:

picking an architecture suitable for a problem,

for image recognition problems – using weights trained on ImageNet,
configuring a network to optimize the results (a long, iterative process).
 In all of these, Keras is a gem. Also, it offers an abstract structure which can be easily
converted to other frameworks, if needed (for compatibility, performance or
anything).

PyTorch is an AI system created by Facebook. Its code is accessible on GitHub and at the
present time has more than 22k stars. It has been picking up a great deal of energy since
2017 and is in a relentless reception development.

CNTK

CNTK allows users to easily realize and combine popular model types such as
feed-forward DNNs, convolutional nets (CNNs), and recurrent networks (RNNs/LSTMs).
It implements stochastic gradient descent (SGD, error backpropagation) learning with
automatic differentiation and parallelization across multiple GPUs and servers. CNTK is
available for anyone to try out, under an open-source license.

Auto ML

Out of all the tools and libraries listed above, Auto ML is probably one of the strongest
and a fairly recent addition to the arsenal of tools available at the disposal of a machine
learning engineer.

As described in the introduction, optimizations are of the essence in machine learning

tasks. While the benefits reaped out of them are lucrative, success in determining optimal
hyperparameters is no easy task. This is especially true in the black box like neural
networks wherein determining things that matter becomes more and more difficult as the
depth of the network increases.

Thus we enter a new realm of meta, wherein software helps up build software. AutoML is
a library which is used by many Machine learning engineers to optimize their models.

Apart from the obvious time saved, this can also be extremely useful for someone who
doesn’t have a lot of experience in the field of machine learning and thus lacks the
intuition or past experience to make certain hyperparameter changes by themselves.

OpenNN

Jumping from something that is completely beginner friendly to something meant for
experienced developers, OpenNN offers an arsenal of advanced analytics.

It features a tool, Neural Designer for advanced analytics which provides graphs and tables
to interpret data entries.

H20: Open Source AI Platform

H20 is an open-source deep learning platform. It is an artificial intelligence tool which is

business oriented and help them to make a decision from data and enables the user to draw
insights. There are two open source versions of it: one is standard H2O and other is paid
version Sparkling Water. It can be used for predictive modelling, risk and fraud analysis,
insurance analytics, advertising technology, healthcare and customer intelligence.

Google ML Kit

Google ML Kit, Google’s machine learning beta SDK for mobile developers, is designed
to enable developers to build personalised features on Android and IOS phones.

The kit allows developers to embed machine learning technologies with app-based APIs
running on the device or in the cloud. These include features such as face and text
recognition, barcode scanning, image labelling and more.

Developers are also able to build their own TensorFlow Lite models in cases where the
built-in APIs may not suit the use case.

ChatGPT Tutorial | ChatGPT Explained | What is

ChatGPT ? | Edureka

This Transform the way you interact with technology with ChatGPT training course.

With this, we have come to the end of our Artificial Intelligence Tools & Frameworks blog.
These were some of the tools that serve as a platform for data scientists and engineers to
solve real-life problems which will make the underlying architecture better and more
robust.

Or If you want to learn how to design intuitive interfaces that captivate users? Join our
cutting-edge Human-Computer Interaction (HCI) Course and take your design skills to the
next level!

R NOW
Tensorflow

On the off chance that you are in the realm of Artificial Intelligence, you have most
likely found out about, attempted or executed some type of profound learning
calculation. Is it accurate to say that they are essential? Not constantly. Is it accurate
to say that they are cool when done right? Truly!

The fascinating thing about Tensorflow is that when you compose a program in
Python, you can arrange and keep running on either your CPU or GPU. So you don’t
need to compose at the C++ or CUDA level to keep running on GPUs.

4. Not mentioned

Pandas

A fast, powerful, and flexible open-source data analysis and data manipulation
library.

Languages: Python

Platforms: Linux, macOS X, Windows

+ Comprehensive data manipulation tools;

+ Seamless integration with many other data science tools;

+ Strong community support;

- Memory consumption can be high;

- Some operations can be slower than pure NumPy.

“Pandas attracts with its user-friendly interface. However, it should be noted

that it can have problems when working with big data science projects (because
it loads all the data it works with into RAM). There is a similar technology
called Dask, which can parallelize its work - it’s really interesting to see its
further development.”
NumPy

A library for the Python language, allowing support for large, multi-dimensional
arrays and matrices along with a vast collection of high-level mathematical functions.

Languages: Python

Platforms: Linux, macOS X, Windows

+ Efficient mathematical operations;

+ Broad functionality for numerical tasks;

+ Seamless integration with other libraries;

- Not specifically tailored for machine learning.

“NumPy does all the computational heavy-lifting for working with vectors
and matrices. This is needed both in Pandas (it is built on NumPy), and in
Scikit-Learn (since it is M-library, here the speed of calculations is
important, which is what NumPy gives us).”

“Sci-kit Learn, Pandas, Numpy - they are the core of ML projects: convenient
both for research and for building production-ready systems, they will continue
to be used everywhere and to develop further.”