Learning Curriculum

Part I: Foundations

Key Concepts

• Mean, median, mode, variance, standard deviation, hypothesis testing
• CSV parsing, string-to-float conversion, data I/O
• Min-max normalization, z-score standardization
• Train/test split, k-fold cross-validation

Part II: Evaluation Metrics

Key Concepts

• Accuracy, Precision, Recall, F1-Score, ROC curve, AUC
• Multi-class confusion matrix interpretation
• Mean Absolute Error, Root Mean Squared Error, R-squared
• Random prediction baselines, ZeroR algorithm

Part III: Linear Models

Key Concepts

• OLS, covariance, correlation, Ridge/Lasso regularization
• SGD algorithm, learning rate, convergence
• Sigmoid function, maximum likelihood, binary classification

Part IV: Classic ML Algorithms

Part V: Neural Networks & Unsupervised

Part VI: Ensemble Methods

Bootstrap, bagging, random forests, boosting concepts - understanding how combining weak learners creates strong predictors.

Part VII: Advanced ML (Industry Track)

Part VIII: Deep Learning Track

Key Concepts

• Convolution layers, transfer learning, data augmentation
• Self-attention, transformer blocks, fine-tuning workflow
• AdamW, learning rate schedules, checkpointing, reproducibility

Part IX: Advanced Algorithms (Expert Track)

Part X: HuggingFace + Kaggle Master Track

Part XI: Frontier Algorithms Track

Parts XII-XIII: Research-to-Production

Distributed training, feature stores, online learning, recommender ranking losses, and causal bandits with Thompson sampling.

Parts XIV-XV: Evaluation & Governance

Part XVI: Causal & Rollout Intelligence

Part XVII: Advanced Boosting & Generative Systems

Key Concepts

• Second-order split gain, Hessian weighting, leaf regularization
• Ordered target encoding, high-cardinality categorical handling
• AdamW internals, warmup-cosine scheduling, gradient clipping
• Train-test domain classification, drift feature ranking
• Forward noising schedule, reverse denoising, sampling diagnostics