Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
An introduction to unsupervised learning, specifically focusing on clustering and the k-means algorithm. Unsupervised learning is a type of machine learning where the training signal or correct answer is not provided. Several views on unsupervised clustering, including summarization, density estimation, classification, and pattern discovery. It also covers different types of unsupervised learning, such as association rule mining and data clustering. The document then delves into the k-means algorithm, explaining its basic algorithm and providing an example of its implementation.
Typology: Study notes
1 / 8
This page cannot be seen from the preview
Don't miss anything!
CSI 5v93: Introduction to machine learning, Lecture 23 – p. 1/
class labels •
regression value •
CSI 5v93: Introduction to machine learning, Lecture 23 – p. 3/
no “correct answer” •
goal is different – the concept to learn is not given with the training data •
overfitting and underfitting are still possible, but it’s not as clear when they occur •
identify and group items which are similar into several groups •
identify and divide items which are different into several groups •
summarize the data with several prototypes •
CSI 5v93: Introduction to machine learning, Lecture 23 – p. 7/
identify and group items which are similar into several groups •
identify and divide items which are different into several groups •
summarize the data with several prototypes •
hierarchical bottom-up •
hierarchical top-down •
-means and Gaussian EM)k iterative “flat” clustering (e.g. •
spectral clustering •
density-seeking (e.g. mean-shift) •
CSI 5v93: Introduction to machine learning, Lecture 23 – p. 9/
troids
(^) • • •
(^) • •
Initial Partition
mber 2
- (^) • •
Iteration Number 20
-meansk Demonstration of
CSI 5v93: Introduction to machine learning, Lecture 23 – p. 13/
-means is actually doingk What
=1 i
=1 j
=1 i
=1 j
=1 m
) X, C( F minimum of local only finds a •
may get stuck in poor solutions •
starting with some solution, always finds a solution that is better (or equal) •
because of the last point, it will always terminate •
major points you learned today •
areas not understood or requiring clarification •