Multi-Class Prediction of a Weighted Time-Varying Association Using Artificial Neural Network in Smart Health Environment

Abstract

Computerization has been a need in recent years. As a result, massive amounts of digital data have been accumulated in practically every industry. Data mining approaches have arisen to serve with this rich data and sparse informations. Data-mining is a process that uses computers to extract and examine hidden patterns in data. Users may understand and benefit from this data mining process of knowledge discovery. According to the value associated with each property, different weights are allocated to those attributes. Weights assigned by the doctors are taken into consideration in this investigation. The first technique given is multiclass weighted associative classification with confidence-based rule ranking, while the second strategy is enhanced, genetic-based rule selection presented as a final step is a distributed multiclass weighted associative classifier. The distributed weighted associative classification is utilized as a consequence of the findings that lower the cost of communication while maintaining the benefit of the centralized approach. The multi class classification's accuracy % rises while using weighted associative classification.