Coursework pertaining to CS5560 : Probabilistic Models in Machine Learning offered in Fall 2018
- Probability revision. Exercises from Kevin Murphy, Machine Learning: A Probabilistic Perspective and Sheldon Ross, Introduction to Probability Models.
- Maximum Likelihood Estimation, Covariances and Correlation. Exercises from Kevin Murphy, Machine Learning: A Probabilistic Perspective.
- Implementation of a Multivariate Gaussian Regression model and inference from data. This was the dataset used.
- Bayesian Linear Regression and Maximum Conditional Likelihood Estimation. Exercises from Kevin Murphy, Machine Learning: A Probabilistic Perspective.
- Generative Classification, Logistic Regression, LDA and its variants. Exercises from Kevin Murphy, Machine Learning: A Probabilistic Perspective.
- Implementation of Gaussian Discriminant Analysis and Naive Bayes' Classifier and compare performance with LIBLINEAR on this dataset.
- Bayesian Learning, Posterior calculation. Exercises from Kevin Murphy, Machine Learning: A Probabilistic Perspective.
- More posterior calculation, Bayesian Bayes Classifier, and MAP estimate for the Naive Bayes model. Exercises from Kevin Murphy, Machine Learning: A Probabilistic Perspective.
- Model selection, Cross-validation. Exercises from Kevin Murphy, Machine Learning: A Probabilistic Perspective.
- Implementation of Gaussian Mixture Models using the EM algorithm and by gradient descent over a loss function. Dataset used is this.