Subject Syllabus Hub367+ MCQs5 Topics

Machine Learning

Supervised and unsupervised learning, regression, classification algorithms, decision trees, neural networks, clustering, and model evaluation metrics.

Topic Syllabus & Revision Notes

High-yield concept summaries, formulas, and common exam pitfalls for each topic.

Topic 16 Key Rules

K-Means & Clustering in Unsupervised Learning

Unsupervised Learning focuses on discovering hidden structures and natural patterns in unlabeled datasets without explicit target labels. Clustering is the core task of grouping similar data instances together such that intra-cluster variance is minimized and inter-cluster variance is maximized. K-Means is the standard centroid-based clustering algorithm. It operates iteratively: 1) Select K initial cluster centroids (typically by randomly choosing K data points or using K-Means++), 2) Assign every data point to its nearest centroid based on Euclidean distance, 3) Recompute the centroid of each cluster as the average (mean) coordinates of all assigned points, and 4) Repeat steps 2 and 3 until centroids stabilize and convergence is achieved. Because 'K' must be specified beforehand, domain experts use techniques like the Elbow Method (plotting WCSS/Inertia vs. K) or Silhouette Analysis to determine the optimal cluster count. Feature scaling (Z-score standardization) is mandatory prior to running K-Means to prevent variables with larger numerical ranges from dominating distance metrics.

View Cheat Sheet & MCQs
Topic 210 Key Rules

Machine Learning & Supervised Learning Fundamentals

Machine Learning algorithms learn predictive patterns from data without being explicitly programmed. Supervised Learning handles labeled target outputs via Regression (predicting continuous numerical quantities) and Classification (predicting categorical labels). Core linear models include Linear Regression (optimized via MSE) and Logistic Regression (using log-loss and sigmoid activation for probability estimation). Model generalization is governed by the Bias-Variance Tradeoff: underfitting stems from insufficient model complexity (high bias), while overfitting arises from excessive model capacity capturing sample noise (high variance). Overfitting is mitigated using Regularization (L1 Lasso for sparse feature selection, L2 Ridge for weight decay), Cross-Validation, Dropout, Early Stopping, and Ensemble Methods (Random Forests, Gradient Boosted Trees). Model performance must be evaluated using domain-appropriate metrics: Accuracy can be misleading on imbalanced datasets, requiring Precision, Recall, F1-Score, and ROC-AUC curves. Feature scaling (standardization/normalization) is mandatory for distance-based models (KNN, SVM) and gradient-based solvers to ensure stable and unbiased convergence.

View Cheat Sheet & MCQs
Topic 3

Recommendation Systems, Artificial Neural Networks, and Deep Learning

Read concise formulas, core definitions, and exam traps.

View Cheat Sheet & MCQs
Topic 4

Semi supervised learning, Reinforcement learning

Read concise formulas, core definitions, and exam traps.

View Cheat Sheet & MCQs
Topic 5

Statistical Learning and Feature Engineering

Read concise formulas, core definitions, and exam traps.

View Cheat Sheet & MCQs