Welcome to the Machine Learning Map of Contents. This dashboard organizes notes logically to provide structured learning paths and rapid discovery.
Foundations
Core vocabulary, problem types, and evaluation, located in the Foundations/ directory.
- Glossary - Foundational machine learning definitions (Bias, Variance, etc.).
- Machine Learning Domains - Breakdown of ML problem types (Classification, Regression, Unsupervised).
- Metrics and Model Evaluation - Guide on evaluating models, bias/variance tradeoff, and regularization.
- Clustering Evaluation Metrics - Purity, Rand Index, Adjusted Rand Index, NMI, and silhouette for judging clusters.
- Loss Functions - How models measure error (MSE, Cross-Entropy).
- Exploratory Data Analysis - The crucial first step of inspecting and understanding your data.
- Scaling Techniques - Methods for normalization and when to use them.
Math for Machine Learning
Mathematical foundations, located in the Math/ directory.
- 0. Glossary (ML Math)
- 1. Probability
- 2. Linear Algebra
- 3. Calculus
- 4. Bayesian Decision Theory
- 5. Determinants
- 6. Fourier Transform
- 7. Chaos Theory
- 8. Law of Large Numbers
- 9. Vector Calculus
- 10. Vector Norms
Classical ML
Deep dives into specific algorithms, located in the Classical ML/ directory.
Linear & Parametric Models
- Linear Classifiers - Hyperplanes and decision boundaries.
- Linear Regression
- Logistic Regression
- Polynomial Regression
- Least-Square Classification
- Fisher’s Linear Discriminant
- Rosenblatt’s Perceptron
- Gradient Descent
Tree-Based & Ensemble Methods
- Decision Trees - Gini splits, pruning, and a worked Iris example with tree visualization.
- Random Forest
- Bagging
- Boosting
Support Vector Machines & Non-Parametric
Bayesian Methods
Unsupervised Learning (Clustering & PCA)
- k-Means - Centroid-based clustering, with a worked Iris example and evaluation.
- Hierarchical Clustering
- Locally Adaptive Clustering (LAC)
- Mixtures of Gaussians
- Principal Component Analysis (PCA)
- Singular Value Decomposition
- Self-organizing Map
Deep Learning
Neural network fundamentals and architectures, located in the Deep Learning/ directory. Sequence models and Transformers live under Natural Language Processing MOC; VAEs and GANs under Generative AI MOC.
- Neural Networks - MLP architecture, backpropagation, and training challenges.
- Activation Functions - Non-linearities and when to use them.
- Batch Normalization - Stable gradients and faster convergence.
- Convolutional Neural Networks - Forward pass, backpropagation through convolutional and pooling layers.
- Autoencoders - Bottleneck architectures, reconstruction loss, and latent spaces.
- Keras vs. PyTorch vs. TensorFlow - Framework comparison.
- Keras Data Generators - Memory-efficient on-the-fly data generation with
keras.utils.Sequence.
Reinforcement Learning
Detailed domain dashboard: Reinforcement Learning MOC
- Reinforcement Learning - The agent-environment interaction loop.
- Markov Decision Processes - Mathematical framework (States, Actions, Transitions).
- Q-Learning - Off-policy value-based learning.
- Deep Reinforcement Learning - Scaling RL with Neural Networks.
- Dynamic Programming - The foundational framework for solving MDPs.
Applications
Located in the Applications/ directory.
- Recommendation Systems
- Learning To Rank
- Anti-Bot Machine Learning - How ML is used to detect and mitigate automated agents.
Related Domains
- Natural Language Processing MOC - Language modeling, embeddings, Transformers, and LLMs.
- Computer Vision MOC - Template matching, filtering, and resolution changes.
- Generative AI MOC - Image generation, VAEs, and GANs.
- Agents MOC - Agent memory, benchmarks, and conversational agents.
To Research / Inbox
(Drop new concepts or terms you want to research here as unlinked wiki-links, e.g., [[Transformers]])