Overview and Objectives

This course is an introduction to the modern AI and ML with equal emphasis on foundational concepts and practice on real world problems.

Course will expose foundations of modern AI along with enough attention to the recent explosion of machine learning techniques such as deep learning. The content of the course is split into three modules which is delivered in the form lectures, hands-on session, case-studies and real world projects.

Contents

Module 1: Problem Formulation and Solving

• Formulating real world problems as AI and ML problems

• Classification and Regression Problems

• Intuitive and Simple Algorithms: KNN, Decsion Tree and Simple Linear Classifier

• Representation of the world and real data: Emphasis on Text, Image, Speech and Sequences

• Visualization, Data Preparation and Unsupervised Learning

• End to end Problem Solving: Navigating through three specific problems and case studies

Module 2: Closer Look at AIML Algorithms

• Simple Linear Algorithms, Optimization and Training

• Non linear Solutions and MLP

• Gradient Descent and Backpropagation

• Decision Tree, Random Forest and Ensembles

• Principles and Practice of ML:

• Training, Validation and Testing
• Overfitting and Regularization
• Errors, Performance Metrics and Reliable Error Estimates

• Support Vector Machines and Kernels

Module 3: Deep Learning and Practical Issues

• Introduction to DL and Toolchain

• Convolutional Neural Networks

• Auto Encoders

• Recurrent Neural Networks

• Selected Special Topics

• Human In the Loop Solutions and Deployment

Components

This course gives equal emphasis on (i) Lectures/classroom learning and (ii) Laboratory sessions and Practice on real-life problems.

Details

• Lectures and classroom learning
• Laboratory experiments with empashsis on foundational algorithms.
• Case studies and Mini-Projects to appreciate the practical utility and nuances.