Module 1 – Advanced Machine Learning (6 hours) 

• Review of Supervised and Unsupervised Learning
• Ensemble methods (Random Forests, Gradient Boosting, XGBoost)
• Dimensionality Reduction (PCA, LDA)
• Model tuning and cross-validation techniques
• Practical:
  • Build classification and regression models using ensemble algorithms
  • Apply PCA on high-dimensional data

Module 2 – Deep Learning and Computer Vision (6 hours) 

• Advanced CNN architecture
• Transfer learning using pre-trained models (VGG, ResNet)
• Introduction to Object Detection and Image Segmentation
• Practical:
  • Transfer learning on a custom dataset

Module 3 – Natural Language Processing with Deep Learning (5 hours) 

• Text mining with TF-IDF
• Advanced word embeddings (Word2Vec, GloVe, BERT)
• Sequence models: LSTM and GRU
• Text generation and summarization
• Transformer architecture overview
• Practical:
  • Sentiment classification using LSTM
  • Text summarization using transformer-based models

Module 4 – Generative AI and Large Language Models (5 hours)

• Fundamentals of Generative Models (GANs, VAEs)
• Introduction to Large Language Models (GPT, Claude, Gemini)
• Prompt Engineering and Fine-tuning basics
• Retrieval-Augmented Generation (RAG)
• Practical:
  • Build a simple image generator using GANs
  • Implement a RAG-based chatbot with Hugging Face or LangChain

Module 5 – Agentic AI and Autonomous Systems (4 hours)

• Introduction to Agentic AI concepts
• Reasoning, memory, and planning in autonomous AI agents
• Tools and frameworks: LangGraph, CrewAI, AutoGen
• Applications: autonomous research assistants, task-oriented agents, digital employees
• Practical:
  •  Build an AI agent capable of multi-step reasoning using LangChain or CrewAI
  •  Implement an autonomous task agent integrating multiple APIs

Module 6 – Reinforcement Learning and Agents (4 hours) 

• RL concepts: Agent, Environment, Reward
• Q-learning and Deep Q Networks (DQN)
• Policy gradients and applications
• AI in gaming and robotics
• Practical:
  •  Implement Q-learning in Python
  • Train an agent to play a simple game

Module 7 – Explainable and Responsible AI (4 hours)

• Explainability techniques (LIME, SHAP)
•  Fairness, Transparency, and Accountability in AI
• Responsible AI frameworks (Google, Microsoft, OECD)
• AI regulation and governance
• Practical:
  • Interpret model predictions using SHAP
  • Case study on bias detection and mitigation

Module 8 – AI System Deployment (3 hours)

• Model deployment using Flask/FastAPI
• Introduction to MLOps (CI/CD, versioning, pipelines)
• Cloud AI platforms (AWS, Azure, GCP)
• Practical:
  • Deploy a trained ML model as an API
  • Automate a simple ML pipeline

Module 9 – Capstone Project & Emerging Trends (3 hours)

• Students identify and solve a real-world problem using AI
• Integration of multiple AI techniques (vision, NLP, prediction, agentic behavior)
•  Presentation and evaluation
• Practical:
  • End-to-end project integrating multiple AI methods
  •  Presentation and viva evaluation