
As an Amazon Associate and affiliate partner, Menrva Books earns from qualifying purchases. Learn more
The goal of clustering is to determine the intrinsic grouping in a set of unlabeled data. But now to decide what constitutes a good clustering? It can be shown that there is no absolute "best" criterion, which would be independent of the final aim of the clustering. Consequently, it is the user, which must supply this criterion, in such a way that the result of the clustering will suit their needs. For instance, we could be interested in finding representatives for homogeneous groups (data reduction), in finding "natural clusters" and describe their unknown properties, in finding useful and suitable groupings or in finding unusual data objects (outlier detection). By grouping the objects in a cluster, a significant improvement in search and in further analysis can be obtained. The following general example illustrates the idea.
This book investigates the fundamental challenge of determining optimal groupings within unlabeled image data for effective segmentation. The author explores the subjective nature of evaluating clustering results, emphasizing that the 'best' criterion is contingent on the user's specific objectives, whether for data reduction, identifying natural clusters, or detecting outliers. The text details how grouping objects within clusters can significantly enhance subsequent analysis and search operations, providing a general framework for understanding these processes.
The book offers a foundational exploration of clustering algorithms as applied to image segmentation, focusing on the user-defined nature of optimal clustering. It highlights the practical benefits of grouping data for improved analysis and search efficiency. The text appears to be a technical overview rather than a review of specific algorithm performance, suggesting its utility for those seeking to understand the principles behind clustering in this context.
Page Count:
81
Publication Date:
2017-12-05
Publisher:
Independently Published
ISBN-10:
1726789918
ISBN-13:
9781726789912
No comments yet. Be the first to share your thoughts!