Prerequisites Roadmap
This is a structured guide to the exact skills you'll need for the projects. Each section lists the core concepts and the best resources to learn them.
Python & Data Science Foundations
The non-negotiable bedrock for all subsequent topics. Master these first.
Core Concepts
- Python Fundamentals: Data structures (lists, dicts), control flow, and functions.
- Environment Management: Using `pip` and `venv` to handle project dependencies.
- NumPy Arrays: Creating, indexing, and performing vectorized operations.
- Pandas DataFrames: Selecting data with `.loc`/`.iloc`, cleaning missing values, and using `.groupby()`.
- Data Visualization: Creating basic plots (histograms, scatter plots) for EDA.
Python - Full University Course
Harvard CS50's Introduction to Programming with Python.
Watch on YouTubePython Data Science Handbook
The definitive open-source book by Jake VanderPlas for learning NumPy, Pandas, and Matplotlib. Essential for data manipulation.
Read OnlineMatplotlib & Seaborn Tutorials
Official Matplotlib documentation and Seaborn gallery with comprehensive examples for data visualization. Perfect for creating publication-quality plots.
Explore TutorialsMathematics for Machine Learning
The core intuitions that power the algorithms. A conceptual understanding is key.
Core Concepts
- Linear Algebra: Intuition behind vectors, matrices, and dot products.
- Statistics: Measures of central tendency (mean/median) and dispersion (std deviation).
- Probability: Understanding of conditional probability and distributions.
- Calculus: The concept of a derivative/gradient as it relates to model optimization.
Essence of Linear Algebra
3Blue1Brown's iconic series. It provides the core intuitions behind vectors and matrices—don't start ML without it.
Watch on YouTubeStatQuest with Josh Starmer
The best channel for understanding statistics and ML concepts. Complex ideas are made "Bam!" easy with clear, visual explanations.
Explore ChannelEssence of Calculus
3Blue1Brown's calculus series that builds intuition for derivatives and gradients. Essential for understanding optimization in ML algorithms.
Watch SeriesCore Machine Learning Competencies
The practical workflow and foundational algorithms used in most traditional ML projects.
Core Concepts
- The ML Lifecycle: Data splitting, feature scaling, training, and evaluation.
- Bias-Variance Tradeoff: The central challenge in building a good model.
- Model Evaluation: Knowing metrics for regression (MSE) and classification (Accuracy, F1-Score).
- Key Algorithms: Linear/Logistic Regression, Random Forest, and XGBoost.
- Unsupervised Learning: Concepts of clustering and anomaly detection.
- Interpretability: Understanding feature importance with tools like SHAP.
Hands-On Machine Learning
Aurélien Géron's book is the bible for practical ML. It perfectly balances theory with hands-on Scikit-Learn examples.
Find the BookMachine Learning Specialization
Andrew Ng's famous Coursera specialization. It's the best place to build a solid theoretical foundation from the ground up.
View on CourseraScikit-Learn Documentation
The official Scikit-Learn documentation with comprehensive tutorials, examples, and API references. Essential for hands-on implementation.
Explore DocumentationAdvanced & Specialized Skills
Domain-specific knowledge for the cutting-edge projects in this repository.
Core Concepts
- Time Series: Understanding trend, seasonality, and time-based features.
- Modern NLP: The role of embeddings and the Transformer architecture.
- LLM Applications (RAG): The pattern of retrieval, prompt engineering, and generation.
- LLM Tooling: Using libraries like LangChain to build pipelines.
- Reinforcement Learning: The core concepts of Agent, Environment, and Reward.
The Hugging Face NLP Course
The official, free course on modern NLP. It's the definitive resource for learning about the Transformer architecture and its ecosystem.
Start the CourseLangChain for LLM Development
DeepLearning.AI's practical course on building LLM applications with LangChain, covering chains, agents, and RAG systems.
View Short CourseFast.ai Practical Deep Learning
Jeremy Howard's revolutionary approach to deep learning. Covers modern techniques including transformers, time series, and practical deployment strategies.
Start Course