Summary
In this chapter, we presented some of the most common soft clustering approaches, focusing on their properties and features. Fuzzy c-means is an extension of the classic k-means algorithm, based on the concept of a fuzzy set. A cluster is not considered a mutually exclusive partition, but rather a flexible set that can overlap some of the other clusters. All of the samples are always assigned to all of the clusters, but a weight vector determines the membership level with respect to each of them. Contiguous clusters can define partially overlapped properties; hence, a given sample can have a not-null weight for two or more clusters. The magnitude determines how much it belongs to every segment.
Gaussian mixture is a generative process ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access