Learning Path

ML Roadmap

A structured learning path from mathematical foundations to production-ready ML knowledge. Follow it in order or jump to any topic you need.

Phase 14/4 available

Zero to ML Foundations

Understand what ML is, how datasets behave, and how the end-to-end pipeline works.

What is Machine Learning?ML Problem Types Dataset Thinking ML Pipeline Overview

Phase 215/15 available

Core ML Thinking

The interview-grade reasoning layer: diagnostics, tradeoffs, and failure analysis.

Overfitting vs Underfitting Bias-Variance Tradeoff Regularization Loss Functions Optimization Generalization Data Leakage Model Complexity Feature Scaling Model Interpretability Error Analysis Model Selection Baseline Models Evaluation Strategy Production Failure Cases

Phase 3

Mathematical Foundations

Build the bedrock before touching algorithms.

Linear Algebra BasicsKhan Academy / 3Blue1Brown

Calculus & DerivativesRequired for gradient descent

Probability & StatisticsMean, variance, distributions

Phase 411/11 available

Supervised Learning

Learn from labeled data — the most common ML paradigm.

Linear Regression Logistic Regression Gradient Descent Regularization (Ridge / Lasso)K-Nearest Neighbors Decision Trees Random Forest Gradient Boosting Support Vector Machine Naive Bayes Gaussian Mixture Models

Phase 59/9 available

Unsupervised Learning

Find structure in unlabeled data.

K-Means Clustering K-Medoid Hierarchical Clustering DBSCAN OPTICS BIRCH Affinity Propagation Mean Shift PCA

Phase 64/7 available

Evaluation & Best Practices

Know when your model is actually good — and how to build it right.

Evaluation Metrics Cross-Validation Feature Engineering

Handling Imbalanced DataComing soon

Missing Data & ImputationComing soon

Hyperparameter Tuning

Anomaly DetectionComing soon

Phase 71/3 available

Advanced Supervised Learning

Push accuracy further with ensemble techniques and probabilistic foundations.

Bias-Variance Tradeoff

Ensemble Methods (Stacking & Blending)Coming soon

Maximum Likelihood EstimationComing soon

Phase 81/3 available

Dimensionality Reduction (Advanced)

Visualize and compress high-dimensional data.

t-SNE

UMAPComing soon

LDA (Linear Discriminant Analysis)Coming soon

Phase 9

Neural Networks

The foundation of modern deep learning.

Neural Network BasicsComing soon

BackpropagationComing soon

Activation FunctionsComing soon