Machine Learning with Mathematica

This course provides the foundational concepts in Machine Learning and Artificial Intelligence.

It teaches about rapid prototyping of AI/ML models for solving tasks of various complexities with Mathematica.

An 8 stage workflow for continuous development and deployment of AI/ML models is covered in detail.

Course Content

Part 1

Machine Learning Overview

• Introduction to general concepts in Machine Learning
• Demonstrations of ML/AI functions in Mathematica

Machine Learning Pipeline

• 8 stage workflow for ML (Data Acquisition, Cleaning, Visualization, Transformation, Training, Cross-validation, Testing, Deployment)
• Developing an end-to-end ML model

Machine Learning Models

• Explanation of working of different models like linear regression, logistic regression, decision trees, random forests, etc.
• Deep dive into optimal training strategies (SGD, ADAM, etc.)
• Model parameters fine tuning methods (Grid search v/s Stochastic search)

Part 2

Neural Networks

• Introduction to neural networks
• Demonstration of simple neural networks

Predictive Neural Networks

• Training neural networks for prediction tasks in various domains:
  • Image classification (Fixed dimension data)
  • Text Sentiment analysis (Variable dimension sequence data)

Generative Neural Networks

• Introduction to Autoencoders as the basis for generative models
• Overview of generative models in various domains:
  • Stable diffusion (Image generation)
  • Transformers (Text generation)