INTRO TO GENERATIVE AI · END-SEMESTER

Your complete GenAI exam study companion

Every one of the 20 lectures — from Machine Learning basics to Agentic AI — explained from first principles with worked examples, runnable code, exam tips and 200+ practice questions with solutions. Study only from here and walk into the exam confident.

Start with Lecture 1 → Try the Mock Exam

Full Lectures

Syllabus Topics

200+

Practice Questions

Mock Exam

How to use this site

The exam is MCQ + coding. This site is built so that reading it cover-to-cover is enough to score top marks. Here is the most effective way to revise:

📖

1. Read each lecture in order

Lectures build on each other. Every page opens with an "In this lecture" map, then explains each syllabus topic with intuition, examples and the exact terms examiners use.

💻

2. Study the code blocks and their output

Coding questions reuse these exact patterns. Every code block shows what it prints — learn to predict the output before you read it.

✅

3. Attempt every practice question

Each lecture ends with MCQs and coding questions. Pick your answer first, then press Check Answer — the solution explains the reasoning, not just the result.

⏱️

4. Finish with the Mock Exam

The full mock exam mixes all 20 lectures, auto-grades your MCQs and reveals every coding solution. Aim for 85%+ before exam day.

💡 Exam-day strategy For MCQs, read every option before choosing — examiners love "all of the above" and near-miss distractors. For coding questions, write the import lines first, then the logic; partial code still earns partial marks. Manage time: roughly 1 minute per MCQ, leave the longest coding question for last.

Exam Toolkit

Four focused study aids built on top of the lectures — use them in the final days of revision.

≡

Cheat Sheet

Every formula, key fact and decision rule for all 20 lectures, condensed onto one print-friendly page.

▦

Interactive Flashcards

120 flip cards for active recall — shuffle, filter by unit and track what you've mastered.

✦

Exam Tips & Traps

An MCQ + coding playbook, plus 18 examiner traps and the misconceptions that quietly cost marks.

</>

Code Patterns Library

Every reusable Python template — sklearn, Keras, NLP, RAG, agents — in one quick-copy reference.

The 20 lectures

Unit 1 · Machine Learning Foundations

Introduction to ML

ML vs traditional programming, AI/ML/DL hierarchy, the four learning types and core ML tasks.

SupervisedUnsupervisedClassification

Data Preprocessing

Missing values, outliers, feature scaling and encoding categorical data the right way.

ImputationScalingEncoding

Linear Regression

Model formulation, OLS, the seven assumptions and the limitations of a straight line.

OLSAssumptions

Logistic Regression

The sigmoid function, log-odds, decision thresholds and why classification needs an S-curve.

SigmoidClassification

Model Evaluation

Confusion matrix, precision/recall, F1, regression metrics, bias-variance and cross-validation.

F1RMSEBias-Variance

Decision Trees

Splitting, Gini & Information Gain, regression trees, overfitting and hyperparameter tuning.

GiniInfo Gain

Unit 2 · Neural Networks & NLP

Intro to Neural Networks

The perceptron, single-layer networks, gradient descent and backpropagation.

PerceptronBackprop

Types of Neural Networks

Multi-layer perceptrons, activation functions and recurrent neural networks (RNNs).

MLPReLURNN

Natural Language Processing

Text preprocessing, tokenization, word embeddings and sequential modelling.

TokenizationEmbeddings

LLM Architecture

Transformers, self-attention (Q/K/V), responsible AI and bias mitigation.

TransformerAttention

Unit 3 · Generative AI Core

Generative AI Modalities

Text, image, code, audio and multimodal models, plus ethics in generative AI.

DiffusionMultimodal

GenAI Commercial APIs

OpenAI & Google APIs, API-key management, costs, rate limits and embeddings.

API KeysHTTP 429

Prompt Engineering

Prompt design, zero/one/few-shot, chain-of-thought, role prompting and refinement.

Few-ShotCoT

Fine-Tuning

Fine-tuning vs prompting, when to fine-tune, instruction-tuning and LoRA/PEFT.

Instruction-TuningLoRA

Managing State in Chatbots

Memory challenges, multi-turn strategies, context management and history passing.

MemoryContext

Unit 4 · Applied GenAI & Agents

RAG

Retrieval-Augmented Generation, grounding LLMs, vector databases and embedding models.

ChromaFAISS

Rapid Prototyping Tools

Why HTML/CSS slows you down, Streamlit, Gradio and building ML UIs in pure Python.

StreamlitGradio

Agentic AI — Components

LLM as brain, tools, memory, planning, and closed vs open-source models.

ToolsOllama

Agentic AI — Control Flow

The ReAct framework, state management and LangGraph graphs, nodes & edges.

ReActLangGraph

Low-Code Automation (n8n)

n8n workflows, nodes, credentials, triggers and replicating LangChain logic visually.

n8nTriggers

🎯 Ready to test yourself? Once you have worked through the lectures, take the full Mock Exam — 40 MCQs and 8 coding questions covering the entire syllabus, with instant scoring.

GenAI Exam Prep · Built from 20 lecture decks, worksheets and notebooks. Content is comprehensive and self-contained — covering, but not limited to, the uploaded material.

Made by Abhinav Bajpai