Unsupervised Machine Learning: Clustering Methods PDF
Document Details
Uploaded by vtbionic
Tags
Summary
This document provides an overview of unsupervised machine learning, focusing on clustering methods. It covers K-Means clustering, explaining the intuition behind the algorithm and demonstrates visual and practical examples. Hierarchical clustering is also discussed, and the document touches on model complexity, bias, and variance.
Full Transcript
Unsupervised Machine Learning Clustering 1 Overview Types of machine learning Unsupervised learning K Means Clustering The intuition A visual demo A practical demo Hierarchical Clustering Model Complexity...
Unsupervised Machine Learning Clustering 1 Overview Types of machine learning Unsupervised learning K Means Clustering The intuition A visual demo A practical demo Hierarchical Clustering Model Complexity 2 3 Supervised vs Unsupervised Learning Example from kNN 4 Supervised vs Unsupervised Learning 5 Overview Types of Machine Learning Unsupervised Learning K Means Clustering The intuition A visual demo A practical demo Hierarchical Clustering 6 K Means Clustering (The Intuition) Y X 7 K Means Clustering (The Intuition) First step: Start with K centroids by putting them in random place Here k = 2 Y X 8 K Means Clustering (The Intuition) Second step: Compute distance of every point from centroid and cluster them accordingly Y X 9 K Means Clustering (The Intuition) Third step: Adjust centroid so that they become center of gravity for given cluster Y X 10 K Means Clustering (The Intuition) Fourth step: Again, re-cluster every point based on their distance with centroid Y X 11 K Means Clustering (The Intuition) Fifth step: Again, Adjust centroids Y X 12 K Means Clustering (The Intuition) Sixth step: Recompute clusters and repeat this till data points stop changing clusters Y X 13 K Means Clustering (The Intuition) Y X 14 How to determine The correct number of clusters (k)? 15 K Means Clustering (How to determine K?) 16 K Means Clustering (How to determine K?) 17 K Means Clustering (How to determine K?) 18 K Means Clustering (How to determine K?) 19 K Means Clustering (How to determine K?) 20 K Means Clustering (How to determine K?) 21 Overview Types of machine learning Unsupervised learning K Means Clustering The intuition A visual demo: https://www.naftaliharris.com/blog/visualizing-k-means-clustering/ A practical demo Hierarchical Clustering 22 Overview Types of machine learning Unsupervised learning K Means Clustering The intuition A visual demo: https://www.naftaliharris.com/blog/visualizing-k-means-clustering/ A practical demo Hierarchical Clustering 23 Hierarchical Clustering 24 Hierarchical Clustering (Cont..) 25 Hierarchical Clustering (Cont..) First step: Assign a cluster to each individual point 26 Hierarchical Clustering (Cont..) Measuring the distance between two points 27 Hierarchical Clustering (Cont..) Measuring the distance between two clusters 28 Hierarchical Clustering (Cont..) The distance matrix 29 Hierarchical Clustering (Cont..) 30 Hierarchical Clustering (Cont..) 31 Hierarchical Clustering (Cont..) 32 Hierarchical Clustering (Cont..) 33 Hierarchical Clustering (Cont..) 34 Hierarchical Clustering (Cont..) 35 Hierarchical Clustering (Cont..) Finished dendrogram 36 Hierarchical Clustering (Cont..) Agglomerative and divisive clustering 37 How to determine cluster # 38 How to determine cluster # 39 How to determine cluster # 40 Model complexity Bias and Variance Bias and Variance (Cont.…) Bias and Variance (Cont.…) Bias and Variance (Cont.…) Bias and Variance (Cont.…) Bias and Variance (Cont.…)
