Machine Learning

Contains coursework assignments in Masters made in latex. Includes solved numericals, understanding questions and some extra topics.

• A1.1 - Machine Learning and its applications
• A1.2 - Least Mean Square (LMS) algorithm
• A1.3 - Confusion matrix and metrics
• A1.4 - Learning system for a Tic-Tac-Toe player
• A2.1 - Match the followiing algorithms and loss functions to their classification counterparts
• A2.2 - The Bias-Variance tradeoff
• A2.3 - Categorical and Numerical features in a dataset
• A2.4 - Maximum Likelihood Estimates (MLE) for the Univariate Gaussian Distribution
• A3.1 - Concept learning and related disciplines
• A3.2 - Use case of concept learning Addison's disease
• A3.3 - Find-S algorithm and Candidate-Elimination algorithm
• A3.4 - Cross validation as an classifier evaluation technique
• A4.1 - Concept learning for Decision Trees
• A4.2 - Decision Tree basics for Machine Learning
• A4.3 - Feature selection and challenges for Decision trees ( use case )
• A4.4 - Iterative Dichotomiser-3 ID-3 algorithm
• A5.1 - Overfitting in Decision Trees with relation to Bias & Variance
• A5.2 - Tree pruning for decision trees ( Reduced Error Pruning )
• A5.3 - Gain ratio as split measure
• A5.4 - Regression Trees
• A6.1 - Perceptron for classification
• A6.2 - The Perceptron training rule ( Delta rule )
• A6.3 - Neural Networks and its modalities
• A6.4 - Activation functions for Neural Networks ( ReLU, Leaky ReLU variants )
• A7.1 - Gradient descent training rule
• A7.2 - Proper loss functions for activation functions
• A7.3 - The Backpropogation algorithm Video
• A7.4 - Effect of Learning rate as hyperparameter
• A8.1 - Non-sequential data classifiers, Feed-forward Neural Networks, BPTT, LSTM
• A8.2 - Naive bayes and Maximum-Aposteriori-Hypothesis (MAP)
• A8.3 - Naive Bayes ( Numerical )
• A8.4 - Spam classification SpamAssassin
• A9.1 - The k-Nearest Neighbor algorithm
• A9.2 - Regression & Classification algorithms
• A9.3 - k-NN ( Numerical )
• A9.4 - Active Learning for Case-based reasoning
• A10.1 - Supervised vs. Unsupervised learning
• A10.2 - k Means algorithm in action
• A10.3 - Hierachical Agglomerative Clustering algorithm
• A10.4 - Fuzzy-C-Means algorithm
• A11.1 - Learning Vector Quantization (LVQ) algorithm
• A11.2 - Reinforcement Learning and its components
• A11.3 - The Value-Iteration algorithm
• A11.4 - The Value-Iteration algorithm ( Episodic process )
• A12.1 - Association rules
• A12.2 - Frequent Itemset Mining ( Exercise )
• A12.3 - Support, Confidence measures for Arules ( Numerical )
• A12.4 - Apriori vs. ECLAT