An Efficient Approach for Detection of Failure in Data Analysis by Using Proposed Modified K-Means Clustering

Authors

  • Ms. Sonia Yadav, Dr. Sachin Sharma

DOI:

https://doi.org/10.17762/msea.v72i1.1946

Abstract

The advent of modern scientific data collection techniques has led to the accumulation of large amounts of information in various fields. Traditional database reference methods are not sufficient to extract useful information from large amounts of data. Cluster analysis is one of the main methods of data analysis, and the k-mean cluster algorithm is widely used in many practical applications. However, the first k-mean algorithm is expensive to calculate, and the quality of the resulting clusters depends largely on the choice of the original centroid. Clustering is an uncontrolled data acquisition (machine learning) technique used to insert data elements into relevant groups without prior knowledge of the group definition. One of the most common and widely studied grouping methods that reduces the error in grouping points in Euclidean space is the K-mean grouping. However, it is known that the k-mean method approaches one of the many local minimums and that the final result depends on the starting points (tools). In this study, we introduced an algorithm for starting a k-tool using the appropriate starting points (tools). Sufficient starting points allow the k-tool to be brought closer to the local minimum; The number of iterations in all data sets is reduced.

Downloads

Published

2023-02-06

How to Cite

Ms. Sonia Yadav, Dr. Sachin Sharma. (2023). An Efficient Approach for Detection of Failure in Data Analysis by Using Proposed Modified K-Means Clustering. Mathematical Statistician and Engineering Applications, 72(1), 593–599. https://doi.org/10.17762/msea.v72i1.1946

Issue

Section

Articles