Advanced_Machine_Learning

This repository consists of notebooks being part of final assignment of Advanced Machine Learning course focused on theory and applicaion of traditional Machine Learning techniques. Each lab was entirely solved by me. Folder data consists of data used in all labs.

Lab 1 - Regression

This lab required:

• implementing Linear Regression class which is able to find solution using explicit formula as well as gradient descent approach
• implementing Ridge Regression class which is able to find solution using explicit formula as well as gradient descent approach
• implementing Lasso Regression class
• implementing Roboust regression class with huber and bisquare weights

Lab 2 - Clustering

This lab required:

• implementing Aglomerative Single Linkage clustering class and analysis of its performance of toy dataset
• implementing Aglomerative Ward Linkage clustering class and analysis of its performance of toy dataset
• Finetuning DBSCAN and HDBSCAN algorithm on given data
• Applying clustering algorithms on real images

Lab 3 - Classification

This lab required:

• Finetuning Logistic Regression, LDA and SVM
• Implementing Linear Discriminative Algorithm class
• Applying above mentioned approaches on toy dataset focused on predicting whether the client subscribed a term deposit

Lab 6 - Model Order Selection

• Applying AIC and BIC criterion to find the best number of mixtures in Gaussian Mixture model.
• Applying AIC and BIC while choosing a model on wine dataset and MNIST dataset
• Estimating accuracy using CLT and Hoeffding inequality