GateShala
Playground

Machine Learning

Machine Learning is the heart of GATE DA. These lessons explain every supervised and unsupervised method in the syllabus from first principles, and implement each one in a few lines of plain Python (no libraries), so you can see exactly how regression weights, tree splits, cluster assignments and principal components are computed.

Start the first lesson

1Foundations

  1. Introduction to Machine LearningWhat machine learning is, supervised and unsupervised learning, regression and classification, features and labels, training and test data, and loss functions, as covered in the GATE DA syllabus.

2Supervised Learning

  1. Linear Regression and Ridge RegressionSimple and multiple linear regression, least squares and the normal equations, R², gradient descent, and ridge regression (L2 regularization) with worked examples for GATE DA.
  2. Logistic RegressionLogistic regression for binary classification, the sigmoid function, the decision boundary, log loss (cross-entropy), gradient descent, odds and log-odds, and softmax, with GATE DA practice questions.
  3. k-Nearest Neighbours and Naive BayesThe k-NN classifier with Euclidean and Manhattan distance, choosing k, feature scaling, and the naive Bayes classifier with Laplace smoothing and Gaussian naive Bayes, with worked GATE DA examples.
  4. Linear Discriminant Analysis and Support Vector MachinesLinear discriminant analysis (LDA) and Fisher's discriminant, the effect of class priors, support vector machines, margins and support vectors, hard and soft margins, hinge loss and the kernel trick, for GATE DA.
  5. Decision TreesHow decision trees classify, entropy and information gain, the Gini index, building a tree with ID3, regression trees, overfitting and pruning, with worked GATE DA examples.
  6. Neural NetworksThe perceptron and its learning rule, why XOR needs a hidden layer, multi-layer perceptrons and feed-forward networks, activation functions, counting parameters, the forward pass and backpropagation, for GATE DA.

3Model Selection

  1. Bias-Variance Trade-off, Cross-Validation and EvaluationThe bias-variance decomposition, overfitting and underfitting, k-fold and leave-one-out cross-validation, the confusion matrix, accuracy, precision, recall, F1 score and ROC curves, for GATE DA.

4Unsupervised Learning

  1. Clustering - k-Means, k-Medoids and Hierarchical ClusteringThe k-means algorithm and its objective, initialization and choosing k, k-medoids, and agglomerative hierarchical clustering with single, complete and average linkage and dendrograms, with worked GATE DA examples.
  2. Dimensionality Reduction and PCAWhy reduce dimensions, principal component analysis step by step, the covariance matrix, eigenvalues and explained variance, PCA through the SVD, and PCA versus LDA, with worked GATE DA examples.

5GATE Practice

  1. Machine Learning GATE Practice SetMixed GATE DA style questions on regression, classification, k-NN, naive Bayes, SVM, decision trees, evaluation, neural networks, clustering and PCA, with answers and explanations.