Executive Certification in Advanced Data Science & Applications

Overview of the Course

• Overview of AI and ML
• The Mathematics required for AI and ML

Linear Algebra for AI

• Linear equations and solutions
• Vectors, Matrices and their Properties
• Inner Products and Norms
• Projections
• Eigenvalues and Eigenvectors
• Singular Value Decomposition

Probability and Statistics for AI

• Probability theory and axioms
• Random variables
• Probability distributions and density functions
• Expectations and moments
• Covariance and correlation
• Hypothesis testing
• MLE, MAP and Bayesian methods

Optimization for Data Science

• Multivariable Calculus
• Unconstrained optimization
• Introduction to least squares optimization
• Gradient based methods

• Overview Python
• Setting up Python environment

• Variables, Data Types
• Control flow (if-else statements, loops)
• Functions and Modules

• List, Tuples
• Dictionaries, Sets
• Classes, Objects and Methods

• NumPy
• Pandas
• Scikit-learn library for ML
• Matplotlib for plotting and visualizations

• PyTorch
• TensorFlow and Keras

• Foundations of Machine Learning – The Machine Learning Paradigm
• Linear and Polynomial Regression
• K-Nearest Neighbors
• Linear Classification – Logistic Regression
• Bias Variance tradeoff,Regularization
• Evaluation methods

• Recap of Linear and Logistic Regression
• Multiclass Classification
• Artificial Neural Networks
  • Artificial Neuron
  • Multilayer Perceptron
  • Universal Approximation Theorem
  • Backpropagation in MLPs
  • Backprop on general graphs
• Optimization in Neural Networks
  • Gradient Descent and its Variants
  • Momentum, Adam, etc.
  • Batch Normalization and other techniques in modern Deep Learning
• Basics of Hyper parameter optimization
• Convolutional Neural Networks (CNN)
  • Introduction
  • CNN Operations
  • CNN Training
  • Image Recognition-SoTA model(s)
  • Object detection/localization - SoTA model(s)
  • Semantic segmentation - SoTA model(s)
• Sequence Analysis Models
  • Recurrent Neural Networks (RNNs)
  • Long Short Term Memory Networks (LSTMs)

• Introduction to Generative Models and their role in Modern AI
• Generative Adversarial Networks (GANs)
• Diffusion Models for image generation
• Transformer Architectures
• Large Language Models (such as ChatGPT)
  • Encoder only models
  • Decoder only models
• Applications of existing Generative AI models
  • Leveraging tools such as ChatGPT in your context
• Future Trends in Generative AI