Unsupervised Learning lessons

This repo contains two lesson tracks I made on key topics in unsupervised learning: Dimensionality Reduction and Clustering. Each track dives deep into the theory, mathematical foundations, and practical implementation using scikit-learn, numpy, pandas and scipy.

Lesson Tracks Overview

1. Dimensionality Reduction

In this track, we explore various matrix factorization-based methods for reducing the dimensionality of data. The lessons include:

• Theory and Mathematics: Understand the mathematical concepts behind dimensionality reduction techniques based on matrix factorization and neighborhood graphs.
• Principal Component Analysis (PCA): Learn how PCA, Sparse PCA, Kernel PCA reduce dimensionality while retaining variance and how to apply it using scikit-learn.
• Non-Negative Matrix Factorization (NMF): Dive into NMF, a powerful technique for extracting meaningful features from non-negative data.
• Practical Implementation: Hands-on tutorials using numpy, scipy, and scikit-learn to apply these techniques to real-world datasets.

2. Clustering

This track focuses on the theory and implementation of various clustering algorithms. The lessons cover:

• Clustering Theory: An overview of clustering concepts, including different clustering paradigms and how to evaluate clustering performance.
• k-Means Clustering: Learn the k-Means algorithm, including its assumptions, limitations, and practical usage.
• Hierarchical Clustering: Explore agglomerative and divisive hierarchical clustering methods.
• DBSCAN: Understand the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm for discovering clusters in spatial data.
• Other Clustering Algorithms: A look at other clustering methods available in scikit-learn, such as Mean-Shift and Spectral Clustering.
• Implementation in Python: Tutorials on implementing and applying these clustering algorithms using scikit-learn.