🤖 AI & Machine Learning
Supervised Machine Learning: Regression and Classification
Master the end-to-end craft of building predictive models for numeric and categorical outcomes. This course spans the fu...
404
Views
Sections
1. Foundations of Supervised Learning0 views
Core concepts, goals, trade-offs, and terminology that underpin regression and classification.
15 topics (15 versions)
1.1Supervised vs Unsupervised vs Reinforcement
1.2Inputs, Targets, and Hypothesis Space
1.3Bias–Variance Trade-off
1.4Underfitting and Overfitting
1.5Empirical Risk Minimization
1.6Loss Functions Overview
1.7Probabilistic Perspective of Supervised Learning
1.8Optimization Basics for ML
1.9Gradient Descent and Variants
1.10Stochasticity and Mini-batching
1.11Evaluation vs Training Objectives
1.12Data Leakage Pitfalls
1.13Reproducibility and Random Seeds
1.14Problem Framing: Regression vs Classification
1.15Types of Supervision and Labels
2. Data Wrangling and Feature Engineering2 views
Practical techniques to clean, encode, scale, and construct informative features while avoiding leakage.
15 topics (15 versions)
2.1Data Types and Tidy Structure
2.2Handling Missing Values
2.3Outlier Detection and Treatment
2.4Categorical Encoding Schemes
1
2.5Ordinal vs Nominal Encodings
1
2.6Text Features: Bag-of-Words and TF-IDF
2.7Date and Time Feature Extraction
2.8Scaling and Normalization Techniques
2.9Binning and Discretization
2.10Interaction and Polynomial Features
2.11Target Leakage in Feature Engineering
2.12Feature Creation from Domain Knowledge
2.13Sparse vs Dense Representations
2.14Feature Hashing Basics
2.15Managing High Cardinality
3. Exploratory Data Analysis for Predictive Modeling1 views
EDA methods tailored to supervised tasks to reveal signal, distribution shifts, and modeling risks.
15 topics (15 versions)
3.1Univariate Distributions and Summary Stats
3.2Pairwise Relationships and Correlations
3.3Visualization for Regression Targets
3.4Visualization for Class Imbalance
3.5Detecting Nonlinearity and Heteroscedasticity
3.6Multicollinearity Diagnostics
3.7Train–Test Split Before EDA
1
3.8Stratification Strategies
3.9Leakage-Aware EDA Practices
3.10Robust Scaling Decisions from EDA
3.11Identifying Data Quality Issues
3.12Feature Importance via Baseline Models
3.13Partial Plots for Early Insight
3.14Handling Out-of-Range Values
3.15Data Imputation Strategy Design
4. Train/Validation/Test and Cross-Validation Strategies0 views
Design robust evaluation schemes and prevent leakage with correct resampling and learning curves.
15 topics (15 versions)
4.1Holdout Validation Principles
4.2K-Fold Cross-Validation
4.3Stratified K-Fold for Classification
4.4Grouped and Blocked CV
4.5Time Series Split Strategies
4.6Nested Cross-Validation
4.7Repeated CV and ShuffleSplit
4.8Leakage-Free Preprocessing within CV
4.9Evaluating Variance of Estimates
4.10Confidence Intervals via Bootstrapping
4.11Model Selection vs Model Assessment
4.12Early Stopping with Validation Curves
4.13Learning Curves Interpretation
4.14Data Snooping and Multiple Testing
4.15Cross-Validation for Imbalanced Data
5. Regression I: Linear Models0 views
Build and diagnose linear regression models, understand assumptions, and evaluate predictive performance.
15 topics (15 versions)
5.1Simple Linear Regression Geometry
5.2Multiple Linear Regression Formulation
5.3Assumptions and Diagnostics
5.4Ordinary Least Squares Solution
5.5Gradient Descent for OLS
5.6Heteroscedasticity and Robust Losses
5.7Transformations of Targets and Features
5.8Categorical Variables in Regression
5.9Interaction Terms in Linear Models
5.10Multicollinearity and VIF
5.11Prediction Intervals vs Confidence Intervals
5.12Feature Scaling Effects in OLS
5.13Handling Outliers with Huber and Quantile Loss
5.14Model Interpretation with Coefficients
5.15Baseline and Dummy Regressors
6. Regression II: Regularization and Advanced Techniques0 views
Control complexity and improve generalization using ridge, lasso, elastic net, and specialized regressors.
15 topics (15 versions)
6.1Ridge Regression Fundamentals
6.2Lasso Regression and Sparsity
6.3Elastic Net and Mixing Parameter
6.4Choosing Regularization Strength
6.5Coordinate Descent Algorithms
6.6Cross-Validated Regularization Paths
6.7Polynomial Regression with Regularization
6.8Generalized Additive Models Overview
6.9Quantile Regression Applications
6.10Poisson and Negative Binomial Regression
6.11Robust Regression Techniques
6.12Feature Selection via L1 Penalty
6.13Bayesian Linear Regression Basics
6.14Multitask and Multioutput Regression
6.15Nonlinear Regression with Kernels
7. Classification I: Logistic Regression and Probabilistic View0 views
Model class probabilities with logistic regression and related probabilistic classifiers.
15 topics (15 versions)
7.1Bernoulli and Binomial Likelihood
7.2Link Functions and the Logit
7.3Maximum Likelihood Estimation
7.4Regularized Logistic Regression
7.5Decision Boundaries and Geometry
7.6One-vs-Rest and Multinomial Logistic
7.7Class Probability Estimation
7.8Feature Scaling and Convergence
7.9Interpreting Coefficients and Odds Ratios
7.10Handling Linearly Separable Data
7.11Class Weights and Cost-Sensitive Learning
7.12Baseline and Dummy Classifiers
7.13Naive Bayes Classifiers
7.14Overfitting in Logistic Models
7.15Sparse High-Dimensional Settings
8. Classification II: Thresholding, Calibration, and Metrics0 views
Make cost-aware decisions by selecting thresholds, calibrating probabilities, and using the right metrics.
15 topics (15 versions)
8.1Confusion Matrix Anatomy
8.2Accuracy, Precision, Recall, F1
8.3ROC Curves and AUC
8.4Precision–Recall Curves and AUC-PR
8.5Threshold Selection Strategies
8.6Cost Curves and Expected Utility
8.7Probability Calibration Methods
8.8Brier Score and Log Loss
8.9Multiclass Metrics and Averaging
8.10Ranking Metrics for Imbalanced Data
8.11Top-k and Coverage Metrics
8.12Macro vs Micro vs Weighted Scores
8.13Cumulative Gain and Lift Charts
8.14Calibration Plots and Reliability
8.15Decision Curves and Net Benefit
9. Distance- and Kernel-Based Methods0 views
Leverage neighborhood and kernel ideas with kNN and SVM for nonlinear decision boundaries.
15 topics (15 versions)
9.1k-Nearest Neighbors for Regression
9.2k-Nearest Neighbors for Classification
9.3Distance Metrics and Scaling Effects
9.4Curse of Dimensionality
9.5Efficient Neighbor Search Structures
9.6Kernel Trick Intuition
9.7SVM for Classification
9.8SVM for Regression (SVR)
9.9Linear vs Nonlinear Kernels
9.10Hyperparameters C, gamma, epsilon
9.11Margin Maximization and Slack Variables
9.12Feature Mapping vs Implicit Kernels
9.13Handling Nonseparable Data
9.14Class Weights in SVMs
9.15Probabilistic Outputs for SVM
10. Tree-Based Models and Ensembles0 views
Learn interpretable trees and powerful ensembles like random forests and gradient boosting.
15 topics (15 versions)
10.1Decision Trees for Regression
10.2Decision Trees for Classification
10.3Impurity and Splitting Criteria
10.4Pruning and Regularization of Trees
10.5Handling Missing Values in Trees
10.6Random Forests Essentials
10.7Extremely Randomized Trees
10.8Gradient Boosting Fundamentals
10.9Learning Rate, Depth, and Estimators
10.10XGBoost, LightGBM, and CatBoost
10.11Feature Importance and Permutation
10.12Partial Dependence and ICE with Trees
10.13Handling Imbalanced Data with Ensembles
10.14Calibration of Ensemble Predictions
10.15Stacking and Blending Strategies
11. Handling Real-World Data Issues0 views
Tackle noise, drift, imbalance, and other practical dataset challenges in production-like settings.
15 topics (15 versions)
11.1Noisy Labels and Annotation Quality
11.2Out-of-Distribution Detection
11.3Data Leakage from Temporal Effects
11.4Drift Detection and Adaptation
11.5Rare Events and Positive-Unlabeled Data
11.6High Cardinality Categorical Features
11.7Skewed Targets in Regression
11.8Missing Not at Random Mechanisms
11.9Data Augmentation for Tabular Data
11.10Weak Supervision and Distant Labels
11.11Semi-Supervised Add-ons to Supervised
11.12Privacy-Preserving Feature Engineering
11.13Federated Learning Basics for Supervised
11.14Small Data and High-D Variants
11.15Shortcut Learning and Spurious Correlation
12. Dimensionality Reduction and Feature Selection0 views
Reduce redundancy and highlight signal with supervised and unsupervised techniques.
15 topics (15 versions)
12.1Filter Methods for Feature Selection
12.2Wrapper Methods and RFE
12.3Embedded Methods with Regularization
12.4Mutual Information for Supervised Tasks
12.5Correlation-Based Feature Pruning
12.6Principal Component Analysis
12.7PCA for Preprocessing Pipelines
12.8Sparse PCA and Kernel PCA
12.9Linear Discriminant Analysis
12.10t-SNE and UMAP for Exploration
12.11Autoencoder Features Overview
12.12Variance Thresholding
12.13Stability Selection Techniques
12.14Feature Selection under Imbalance
12.15Interpreting Reduced Dimensions
13. Model Tuning, Pipelines, and Experiment Tracking0 views
Automate workflows, search hyperparameters, and track experiments reproducibly.
15 topics (15 versions)
13.1Grid Search and Random Search
13.2Bayesian Optimization Basics
13.3Successive Halving and Hyperband
13.4Early Stopping and Warm Starts
13.5Hyperparameter Spaces and Priors
13.6Pipeline Composition and Caching
13.7ColumnTransformers for Heterogeneous Data
13.8Custom Transformers and Estimators
13.9Cross-Validated Pipelines
13.10Refit Strategies and Model Persistence
13.11Reproducible Experiment Tracking
13.12Logging and Metadata Management
13.13Parallel and Distributed Tuning
13.14Budget-Aware Optimization
13.15Reusing and Sharing Artifacts
14. Model Interpretability and Responsible AI0 views
Explain model behavior, assess fairness, and communicate uncertainty responsibly.
15 topics (15 versions)
14.1Global vs Local Explanations
14.2Coefficient-Based Interpretation
14.3Permutation Importance Pitfalls
14.4SHAP Values for Trees and Linear Models
14.5LIME for Local Explanations
14.6Counterfactual Explanations
14.7Partial Dependence and ICE Best Practices
14.8Feature Interaction Analysis
14.9Monotonic Constraints in Models
14.10Detecting and Mitigating Bias
14.11Fairness Metrics and Trade-offs
14.12Privacy Risks in Supervised Models
14.13Adversarial Examples in Tabular Data
14.14Transparency and Documentation
14.15Human-in-the-Loop Review
15. Deployment, Monitoring, and Capstone Project0 views
Ship models to production, monitor performance, and complete an end-to-end capstone.
15 topics (15 versions)
15.1Exporting and Serializing Models
15.2Batch vs Real-Time Inference
15.3Feature Stores and Data Contracts
15.4Model Serving Patterns and APIs
15.5Containerization and Reproducibility
15.6Hardware Acceleration Considerations
15.7A/B Testing and Shadow Deployments
15.8Monitoring Performance and Drift
15.9Alerting and Incident Response
15.10Retraining Triggers and Schedules
15.11Model Governance and Compliance
15.12Testing and CI for ML Systems
15.13Secure and Responsible Deployment
15.14Cost Optimization for Inference
15.15Capstone Project Brief and Milestones