Roadmap
Machine Learning Roadmap
My exploration of machine learning paradigms, algorithms, and real-world applications.
| Order | Field | Learn About (Concepts) | Learn By Doing (Hands-on Coding/Projects) | Best Free Resources |
|---|---|---|---|---|
| 1 | Foundations | Python for ML, linear algebra, calculus, probability & statistics, data preprocessing | NumPy/Pandas exercises, exploratory data analysis on public datasets (Iris, Titanic) | Andrew Ng Machine Learning (Coursera) Google ML Crash Course roadmap.sh Machine Learning |
| 2 | Supervised Learning | Regression (linear, logistic), classification (decision trees, SVM, KNN, random forests), evaluation metrics (accuracy, precision, recall, F1, ROC-AUC) | Predict house prices, classify images or spam, build ensemble models with scikit-learn | DeepLearning.AI ML Specialization scikit-learn Official Tutorial Kaggle Intro to ML |
| 3 | Unsupervised Learning | Clustering (K-means, hierarchical, DBSCAN), dimensionality reduction (PCA, t-SNE), association rules, anomaly detection | Customer segmentation, visualize high-dimensional data, compress images with autoencoders | Andrew Ng ML Course (Unsupervised section) scikit-learn Clustering Guide |
| 4 | Deep Learning | Neural networks, backpropagation, activation functions, CNNs, RNNs/LSTMs, Transformers, overfitting & regularization | Build image classifiers (CNN), sequence models, train with PyTorch/TensorFlow/Keras | fast.ai Practical Deep Learning DeepLearning.AI Deep Learning Specialization PyTorch Tutorials |
| 5 | Computer Vision | Image processing, convolutions, object detection (YOLO, SSD), segmentation, transfer learning with pre-trained models | Build image classifier, face detection, object detection app, style transfer | Stanford CS231n (Convolutional Networks) TensorFlow CV Tutorials OpenCV Documentation & Tutorials |
| 6 | Natural Language Processing | Text preprocessing, embeddings (Word2Vec, GloVe), Transformers, sentiment analysis, machine translation, LLMs basics | Sentiment classifier, text generation, question answering, build a simple chatbot | Hugging Face NLP Course DeepLearning.AI NLP Specialization NLTK Book |
| 7 | Reinforcement Learning | Markov decision processes, Q-learning, policy gradients, actor-critic, deep RL (DQN) | Train agent for CartPole, simple games (OpenAI Gym), grid world navigation | OpenAI Spinning Up Sutton & Barto RL Book (free PDF) Gymnasium (formerly OpenAI Gym) |
| 8 | Generative AI & Self-Supervised Learning | GANs, VAEs, diffusion models, self-supervised techniques, contrastive learning | Generate images with GANs or diffusion, masked autoencoders, fine-tune LLMs | Generative AI with Large Language Models (DeepLearning.AI) Hugging Face Diffusers |
| 9 | Advanced Topics | Semi-supervised & transfer learning, ensemble methods, explainable AI (XAI), MLOps basics, ethical considerations | Few-shot learning projects, model interpretation with SHAP/LIME, deploy a model | TensorFlow Extended (TFX) for MLOps Interpretable ML Book |
Disclaimer: This learning path links to third-party websites and free resources. I do not own or control these external sites. Content and availability may change over time. Use at your own risk.