Machine Learning

This repository is for my personal learning and interview preparation.

It should be deep, clear, structured, and useful enough that I can study from it seriously.

Purpose

The purpose of this repository is to make machine learning study concrete. It connects concepts to examples, interview prompts, diagrams, exercises, case studies, and portfolio projects so each topic can be reviewed, practiced, and explained clearly.

Who This Repository Is For

• Software engineers moving into AI or machine learning engineering.
• Data scientists who want stronger production and system design foundations.
• Students who want a clear path from fundamentals to applied projects.
• Practitioners preparing for AI Engineer, Machine Learning Engineer, LLM Engineer, Data Scientist, Applied Scientist, or GenAI Engineer interviews.

Start Here

Start with What Is Machine Learning?. Then read the numbered files in each folder according to the recommended order below. Each learning file includes intuition, formal explanation where useful, real-world relevance, common mistakes, interview angles, an exercise, a Mermaid diagram, and navigation links.

Beginner Path

1. Read fundamentals/, math/, and statistics/.
2. Run the first three notebooks in notebooks/.
3. Complete the fundamentals, linear algebra, and statistics quizzes.
4. Build one small baseline model from src/ml_from_scratch/.

Deep Study Path

Use the full curriculum order from fundamentals through production AI. Take notes on assumptions, metrics, failure modes, and deployment concerns after each folder. Pair every conceptual section with one quiz, one cheatsheet review, and one small implementation or case study.

Interview Preparation Path

Use interview-prep/README.md after completing the core technical sections. Practice explaining tradeoffs out loud, then use mocks/README.md for full-round simulation and scoring.

Project and Case Study Path

Use case-studies/README.md to learn how systems are framed and reviewed. Use capstone-projects/README.md to turn that knowledge into portfolio projects with evaluation, failure analysis, and interview explanation.

Learning Roadmap

flowchart LR
    Basics --> Math --> Statistics --> Python_and_Data[Python and Data]
    Python_and_Data --> Classical_ML[Classical ML] --> Deep_Learning[Deep Learning]
    Deep_Learning --> NLP --> Computer_Vision[Computer Vision] --> Recommenders
    Recommenders --> MLOps --> GenAI --> LLMs
    LLMs --> Vector_Databases[Vector Databases] --> RAG --> Agents
    Agents --> Production_AI[Production AI] --> ML_System_Design[ML System Design]
    ML_System_Design --> Ethics --> Case_Studies[Case Studies] --> Interviews --> Projects

Loading

Table of Contents

Section	What It Contains
Fundamentals	Core machine learning vocabulary, workflows, data splits, generalization, and model evaluation habits.
Math	Linear algebra, calculus, optimization, distance metrics, and information theory for machine learning.
Statistics	Probability, uncertainty, inference, testing, sampling, causality, and experimentation.
Data Science	Practical data cleaning, exploration, feature engineering, visualization, experimentation, and communication.
Classical ML	Supervised and unsupervised machine learning algorithms, evaluation, interpretability, and model selection.
Deep Learning	Neural networks, optimization, regularization, convolution, sequence models, attention, and transformers.
NLP	Text preprocessing, tokenization, embeddings, sequence modeling, transformers, semantic search, and NLP evaluation.
Computer Vision	Image tensors, convolution, classification, detection, segmentation, vision transformers, and multimodal models.
Recommender Systems	Candidate generation, ranking, collaborative filtering, matrix factorization, and recommender evaluation.
MLOps	Reproducibility, experiment tracking, model registries, serving, monitoring, CI/CD, and governance.
Generative AI	Autoregressive models, VAEs, GANs, diffusion, multimodal generation, and generative AI evaluation.
LLMs	Transformer decoder architecture, pretraining, instruction tuning, prompting, tools, evaluation, and serving.
Vector Databases	Embeddings, similarity search, ANN indexes, filtering, hybrid search, reranking, and vector search scaling.
RAG	Document ingestion, chunking, embeddings, retrieval, reranking, generation, evaluation, observability, and security.
Agents	Agent loops, tool use, planning, memory, multi-agent design, evaluation, observability, and risk controls.
Production AI	Architecture patterns, latency, cost, caching, routing, fallbacks, privacy, monitoring, and product metrics.
Machine Learning System Design	System design practice for ML platforms, recommendation systems, search ranking, RAG, agents, evaluation, and real-time inference.
Ethics Safety	Fairness, privacy, security risks, misuse, responsible AI, and governance.
Case Studies	Applied ML and AI system case studies with production tradeoffs and interview discussion points.
Interview Prep	Role-specific roadmaps, technical question sets, behavioral preparation, and final revision checklists.
Mocks	Mock interview rounds with prompts, expected answers, scoring rubrics, and self-review material.
Quizzes	Ten-question review sets with answer keys and explanations for retrieval practice.
Cheatsheets	Compact reference pages for fast revision before projects and interviews.
Capstone Projects	Project guides for portfolio work, implementation planning, evaluation, and interview explanation.
Notebooks	Hands-on starter notebooks for NumPy, pandas, classical ML, neural networks, transformers, embeddings, and RAG.
Source Examples	Runnable educational Python examples for ML from scratch, RAG, agents, and simple serving.
Diagrams	Reusable Mermaid diagram source files for learning paths and architecture sketches.
Tools	Repository maintenance scripts for indexing, link validation, and Markdown quality checks.

Repository Index

Use REPO_INDEX.md when you want a single grouped list of every Markdown file in the repository. Regenerate it after adding, renaming, or removing learning material.

Recommended Reading Order

Fundamentals -> Math -> Statistics -> Data Science -> Classical ML -> Deep Learning -> NLP -> Computer Vision -> Recommender Systems -> MLOps -> Generative AI -> LLMs -> Vector Databases -> RAG -> Agents -> Production AI -> Machine Learning System Design -> Ethics Safety -> Case Studies -> Interview Prep -> Mocks -> Quizzes -> Cheatsheets -> Capstone Projects

Progress Checklist

• Finish fundamentals/ and complete one retrieval practice item.
• Finish math/ and complete one retrieval practice item.
• Finish statistics/ and complete one retrieval practice item.
• Finish data-science/ and complete one retrieval practice item.
• Finish classical-ml/ and complete one retrieval practice item.
• Finish deep-learning/ and complete one retrieval practice item.
• Finish nlp/ and complete one retrieval practice item.
• Finish computer-vision/ and complete one retrieval practice item.
• Finish recommender-systems/ and complete one retrieval practice item.
• Finish mlops/ and complete one retrieval practice item.
• Finish generative-ai/ and complete one retrieval practice item.
• Finish llms/ and complete one retrieval practice item.
• Finish vector-databases/ and complete one retrieval practice item.
• Finish rag/ and complete one retrieval practice item.
• Finish agents/ and complete one retrieval practice item.
• Finish production-ai/ and complete one retrieval practice item.
• Finish machine-learning-system-design/ and complete one retrieval practice item.
• Complete at least five case studies.
• Complete at least three mock interviews.
• Finish one capstone project with a written evaluation report.

Using Quizzes

Use quizzes for retrieval practice, not passive reading. Answer first, then check the key and write a one-sentence correction for every miss.

Using Cheatsheets

Use cheatsheets before interviews and project work. They are compact references for concepts, metrics, traps, and answer structure.

Using Mocks

Use mocks to practice the full interview loop: clarification, baseline, design, evaluation, tradeoffs, and self-review. Score yourself against the rubric and repeat weak rounds.

Using Projects and Case Studies

Case studies show how to reason about real systems. Capstone projects turn that reasoning into implementation practice. For each project, write down the problem, dataset, baseline, metric, error analysis, production risks, and interview explanation.

Repository Tools

• python tools/generate_repo_index.py refreshes REPO_INDEX.md.
• python tools/check_links.py validates local Markdown links.
• python tools/check_markdown_quality.py validates structure, naming, navigation, and formatting.
• python tools/fix_markdown_formatting.py applies safe formatting and navigation normalization.

Repository Scope

This repository is maintained for personal study and interview preparation. It is not structured as a community-maintained open-source project. The root is intentionally kept focused on learning paths, reference material, practice, projects, and validation tools.

Quality Promise

This repository aims to be practical, readable, and honest about tradeoffs. Content should teach from first principles, connect to real engineering work, include practice, and avoid marketing language or unsupported claims.