Clustering Unsupervised learning introduction Machine Learning Supervised learning

Supervised learning Training set: Andrew Ng

Unsupervised learning Training set: Andrew Ng

Applications of clustering Market segmentation Social network analysis Image credit: NASA/JPL-Caltech/E. Churchwell (Univ. of

Clustering Machine Learning K-means algorithm

K-means algorithm Input: (number of clusters) - Training set (drop convention) Andrew Ng

K-means algorithm Randomly initialize cluster centroids Repeat { for = 1 to : =

K-means for non-separated clusters Weight T-shirt sizing Height Andrew Ng

Clustering Optimization objective Machine Learning

K-means optimization objective = index of cluster (1, 2, …, ) to which example

Clustering Random initialization Machine Learning

Random initialization Should have Randomly pick examples. Set examples. training equal to these Andrew

Random initialization For i = 1 to 100 { Randomly initialize K-means. Run K-means.

Clustering Choosing the number of clusters Machine Learning

Choosing the value of K Cost function Elbow method: 1 2 3 4 5

Choosing the value of K Sometimes, you’re running K-means to get clusters to use

Slides: 29

Download presentation

Clustering Unsupervised learning introduction Machine Learning

Supervised learning Training set: Andrew Ng

Unsupervised learning Training set: Andrew Ng

Applications of clustering Market segmentation Social network analysis Image credit: NASA/JPL-Caltech/E. Churchwell (Univ. of Wisconsin, Madison) Organize computing clusters Astronomical data analysis Andrew Ng

Clustering Machine Learning K-means algorithm

Andrew Ng

K-means algorithm Input: (number of clusters) - Training set (drop convention) Andrew Ng

K-means algorithm Randomly initialize cluster centroids Repeat { for = 1 to : = index (from 1 to ) of cluster centroid closest to for = 1 to : = average (mean) of points assigned to cluster } Andrew Ng

K-means for non-separated clusters Weight T-shirt sizing Height Andrew Ng

Clustering Optimization objective Machine Learning

K-means optimization objective = index of cluster (1, 2, …, ) to which example assigned = cluster centroid ( ) = cluster centroid of cluster to which example assigned Optimization objective: is currently has been Andrew Ng

Clustering Random initialization Machine Learning

Random initialization Should have Randomly pick examples. Set examples. training equal to these Andrew Ng

Local optima Andrew Ng

Random initialization For i = 1 to 100 { Randomly initialize K-means. Run K-means. Get Compute cost function (distortion) . } Pick clustering that gave lowest cost Andrew Ng

Clustering Choosing the number of clusters Machine Learning

What is the right value of K? Andrew Ng

Choosing the value of K Cost function Elbow method: 1 2 3 4 5 6 (no. of clusters) 7 8 1 2 3 4 5 6 7 8 (no. of clusters) Andrew Ng

Choosing the value of K Sometimes, you’re running K-means to get clusters to use for some later/downstream purpose. Evaluate K-means based on a metric for how well it performs for that later purpose. E. g. T-shirt sizing Weight T-shirt sizing Height Andrew Ng