The proposed intelligent mining framework for highly dense statistics clustering in healthcare networks is a unique approach that leverages the electricity of the country of the artwork machine studying algorithms to plot a practical solution to the mission of efficient records clustering in healthcare networks. The framework utilizes an ensemble of supervised and unsupervised mastering strategies to extract meaningful insights from excessively dense information clusters using uncovering styles in the statistics and deriving actionable know-how. The proposed framework is based on the fusion of an advanced W-Murkowski distance, a generalization of the Euclidean distance, with a changed BIC (Bayesian Information Criterion). The W-Murkowski and BIC distances are merged to form a hybrid distance metric. It is then applied to the proposed clustering method. Moreover, this method is based upon characteristic scaling strategies, employing scaling of both unmarried attributes and multivariate attributes. Furthermore, the framework carries a weighted vote-casting approach that allows for leveraging the strengths of unsupervised and supervised mastering algorithms for statistics clustering. The weighted voting approach combines the outcomes received from multiple algorithms, inclusive of okay-way and k-Modes, to supply a consensus output that is more dependable than the outcomes received from today’s healthcare surroundings, information safety, and privateness has emerged as a primary concern with the increasing quantity of records generated. With the emergence of recent statistics assets like affected person-logged records, beacons, wearable’s, and clinical scanners, it has become challenging to effectively save, manage, and extract beneficial insights from the massive amount of facts. Conventional clustering algorithms for excessive-density statistics in healthcare networks have shown to be computationally in-depth and inadequate to handle the considerable complexity of healthcare data. Therefore, a want exists for a clever mining framework that can efficaciously deal with the intense stage of high-dense records clusters. The proposed intelligent mining framework provides a green way to high-density clustering issues in healthcare networks. This framework combines several clustering strategies, which include Divide-and-triumph over, Hierarchical Clustering, and Density-based total Clustering, to pick out accurate and significant clusters in a network. The Divide-and-triumph method divides the dataset into smaller subsets for better clustering accuracy. The Hierarchical Clustering set of rules is used to construct a hierarchical structure between the subsets for further clustering.
M. Ramkumar, R. Karthick, A. Jeyashree
Knowledge Institute of Technology, India
Clustering, Construct, Healthcare, Structure, Algorithms, Characteristic
| January | February | March | April | May | June | July | August | September | October | November | December |
| 0 | 3 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Published By :
ICTACT
Published In :
ICTACT Journal on Data Science and Machine Learning
( Volume: 4 , Issue: 4 , Pages: 485 - 490 )
Date of Publication :
September 2023
Page Views :
293
Full Text Views :
11
|