University:
University of South Wales
Course:
G40C | Computer Science
Academic year:

2023
Views:

78

Pages:

16
Author:

Rebecca L

MODULE-6 Clustering It is the task of grouping a set of objects in such a way that objects in the same group are more similar to each other than those in the other groups. * Clustering is always done on an unlabelled set of data. * Unsupervised way of learning. *st deals with finding a structure in the collection of data. Applications of Clustering cis Marketing (ⅱ) Insurance (ii) www (iv) Image segmentation Basic Criteria for Clustering * An important component of a clustering algorithm is the distance measure between the data points.. Majorly used distance measure is the "kundian Distance". Consider two points in a 20 space K-Means Clustering Input: Dataset without labels Number clusters k Output: Dataset divided into to clusters Algorithm: • Iterate until stable (no object move from the group) Step 1: Determine the centroid coordinate Step 2: Determine the distance of each object Step 3: Group the object based on the minimum distance. Flowchart Density-based Clustering In density based clustering, clusters are defined of higher density than the remainder of the dataset." as areas Density + het 'E' be some constant het 'p' be an arbitrary point of Then, & neighbourhood is defined as {q: d(p,q)<&} High Density * Suppose mo is an integer value, we say that a point 'p' high density, if & neighbourhood (p) contains mo points. Core Point * If & neighbourhood (P) contains more than mo points [mo+non-negative value] Border Point * The points which has less number of neighbours (