Advanced AI Course (Level 2)

Duration: 4 Weeks (5 Days per Week) 

Delivery Mode: Hybrid (Theory + Hands-on + Self-Study) 

Daily Commitment: 2 Hours 

Total Duration: 30 Days 

Fees: 34999/-

Tools Required: 

The following tools and platforms will be used throughout the course for development, experimentation, and deployment: 

Programming Languages & Libraries 

• Python 3.10+ 

• NumPy, Pandas, Matplotlib, Seaborn 

• Scikit-learn • PyTorch / TensorFlow (choose based on preference) 

• NLTK, spaCy 

• Transformers (HuggingFace) 

• OpenAI API 

• LangChain, FAISS / ChromaDB (for RAG & Vector DB) Development Environments 

• Google Colab (for cloud-based GPU experiments) 

• Jupyter Notebooks (for local development) 

• VS Code / PyCharm (for advanced editing and debugging) Deployment & Visualization 

• Streamlit (for model UI and API presentation) • Flask / FastAPI (for backend deployment) 

• Git (version control & collaboration) 

• GitHub / GitLab (code hosting & project management) Model Tracking & Experimentation 

• Weights & Biases / MLflow (for model versioning and performance tracking) Explainability & Ethics 

• SHAP, LIME, Captum (for model interpretability) 

Prerequisites (Before Starting the Course) Learners should have: 

Good Python programming skills 

• Functions, classes, list comprehensions, file handling 

• Basic knowledge of Linear Algebra & Calculus 

◦ Matrices, vectors, gradients, derivatives 

• Basic Statistics & Probability 

◦ Mean, variance, standard deviation, conditional probability 

• Machine Learning Foundations 

◦ Supervised & Unsupervised learning 

◦ Familiarity with scikit-learn 

• Basic understanding of Neural Networks Forward/backward propagation, loss functions 

• Machine Learning and Data Analytics

Curriculum:

Week 1: Deep Learning Fundamentals (12 Hours) 

Day 1: Recap of Neural Networks, Activation Functions 

Day 2: Optimization Algorithms – SGD, Adam, RMSprop 

Day 3: Regularization – Dropout, L1/L2, Batch Normalization 

Day 4: Model Evaluation – Precision, Recall, F1, AUC 

Day 5: Convolutional Neural Networks – Convolution, Pooling, Architectures (VGG, ResNet) 

Day 6: Hands-on CNN Implementation using PyTorch/TensorFlow 

Week 2: Natural Language Processing (NLP) (12 Hours) 

Day 7: NLP Basics – Tokenization, Word Embeddings 

Day 8: Word Embedding Techniques – Word2Vec, GloVe, FastText 

Day 9: Sequence Models – RNN, LSTM, GRU 

Day 10: Attention Mechanism & Seq2Seq Models 

Day 11: Transformers – Self-Attention, Encoder-Decoder Architecture 

Day 12: Introduction to BERT & Large Language Models (LLMs) 

Week 3: Generative AI & Foundation Models (12 Hours) 

Day 13: GANs – Architecture, Types (DCGAN, CycleGAN) 

Day 14: Diffusion Models – Fundamentals and Applications 

Day 15: Hands-on GANs – Image Generation 

Day 16: HuggingFace Transformers – Using Pretrained Models 

Day 17: Fine-tuning BERT, GPT for NLP Tasks 

Day 18: Prompt Engineering and Inference Techniques 

Week 4: RL, Explainability, Projects & Tools (12 Hours) 

Day 19: Reinforcement Learning Basics – MDPs, Q-Learning 

Day 20: Deep Q-Networks (DQN), Policy Gradients 

Day 21: Multi-agent RL & Real-World Use Cases 

Day 22: Explainable AI – SHAP, LIME, Captum 

Day 23: Ethical AI – Bias, Fairness, Responsible AI 

Day 24: AI Deployment – APIs, Streamlit 

Week 5: Projects (12 Hours) 

Day 25: Project 1 – Image Classification with Transfer Learning 

Day 26: Project 2 – Text Classification using BERT 

Day 27: Project 3 – Chatbot using Transformers 

Day 28: LangChain & RAG (Retrieval Augmented Generation) 

Day 29: AI Tools – OpenAI API, LLM Agents, Vector DBs 

Day 30: Git, Model Versioning, Collaboration Tools 

Certification 

• Issued by KITS TECH LEARNING CENTER 

• Other certifications are optional and chargeable.