Complete Machine Learning Notes

Overview

This repository documents my entire journey of learning Machine Learning (ML) — from the machine learning algorithms to large language models.

It is designed to serve as a complete reference for anyone looking to understand ML, DL and LLM concepts deeply and systematically.

• Machine Learning Algorithms: Covers fundamental algorithms, their workings, and applications.

• Deep Learning: Explores neural networks, architectures, and training techniques.

• Transformers and Large Language Models: Delves into the architecture, training, and applications of transformers and LLMs.

---

How to Use This Repository

• It is recommended to fork this repository to have your own copy for personal use and future reference.

• You can download the notes as PDF files for offline study as well.

---

Repository Structure

• Introduction to Machine Learning - NPTEL IIT Madras Notes

  • 1. Linear Regression and Regularization Techniques

  • 2. Logistic Regression and Linear Discriminant Analysis

  • 3. Support Vector Machines (SVMs)

  • 4. MLE, MAP and Bayesian Estimation

  • 5. Decision Trees

  • 6. Ensemble Methods ~ Baggging and Boosting

  • 7. Bayesian Networks and Naive Bayes Classifier

  • 8. Clustering Methods ~ KMeans, Hierarchical and DBSCAN

  • 9. BIRCH and CURE Algorithms

  • 10. Gaussian Mixture Models (GMM) and Expectation-Maximization (EM) Algorithm

• Complete Deep Learning Notes

  • 1. Introduction to ANN

  • 2. Forward and Backward Propagation

  • 3. Activation Functions and Cost Functions

  • 4. Optimizers in Deep Learning

  • 5. Weight Initialization techniques

  • 6. Convolutional Neural Networks (CNNs)

  • 7. Recurrent Neural Networks (RNNs)

  • 8. LSTM and GRU Networks

  • 9. Seq2Seq Models and Attention Mechanism

---

Resources Used

Some of the materials and courses referenced:

• Andrew Ng — Machine Learning Specialization

• Andrew Ng - DeepLearning Specialization

• NPTEL — Machine Learning

• NPTEL - Introduction to LLMs

• Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow — Aurélien Géron

---