Scholarlink

<p>With today’s expanding population, healthcare is a major concern. It includes a variety of strategies for enhancing health, including early disease detection, prevention, and treatment. As a result, decision-making involves knowledge extraction at many levels. One way to solve this issue is to infer knowledge from the information system, which results in many stages of knowledge extraction. Another challenging issue is handling uncertainties during data analysis. To deal with these uncertainties, computer intelligence approaches must be employed when choosing features, categorizing and clustering objects, and creating decision rules. Recent research has shown that bio-inspired computing methods can be utilized to classify a system using the best possible feature selection. On the contrary, rough set helps in generating decision rules from an information system. It generates decision rules without loss of information. At the same time, swarm optimization is a powerful method in identifying chief features while solving a real world problem but fails to handle uncertainties. Therefore, to analyze real world information system while handling uncertainties, an integrated approach that integrates rough set and cat swarm optimization is proposed. The proposed method identifies the best features concerning degree of dependency of rough set as fitness function in the cat swarm optimization. Further, decision rules are generated using rough set. The feasibility of the suggested research in comparison to existing research is demonstrated by an empirical investigation on the cirrhosis disease. The proposed model that integrates rough set and cat swarm optimization (RSCSO) is analyzed over the detection of cirrhosis disease. It attains an accuracy of 92.06%. Besides, the proposed model RSCSO is compared with the performance of the decision tree and the conventional rough set model. The accuracy attained for decision tree model is 81.6% whereas that of rough set model is 89.08%. Further, the accuracy of RSCSO is compared with cat swarm optimization - decision tree (CSODT) model. The accuracy attained by CSODT model is 86.4%. The analysis shows that the RSCSO model is more accurate as compared to RS, decision tree and CSODT model.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An integrated cat swarm rough set optimization algorithm for the detection of cirrhosis disease

  • 期刊论文

  • 发表日期：

    2026年2月17日

查看全文
  • Madhusmita Mishra,
  • D. P. Acharjya

摘要

With today’s expanding population, healthcare is a major concern. It includes a variety of strategies for enhancing health, including early disease detection, prevention, and treatment. As a result, decision-making involves knowledge extraction at many levels. One way to solve this issue is to infer knowledge from the information system, which results in many stages of knowledge extraction. Another challenging issue is handling uncertainties during data analysis. To deal with these uncertainties, computer intelligence approaches must be employed when choosing features, categorizing and clustering objects, and creating decision rules. Recent research has shown that bio-inspired computing methods can be utilized to classify a system using the best possible feature selection. On the contrary, rough set helps in generating decision rules from an information system. It generates decision rules without loss of information. At the same time, swarm optimization is a powerful method in identifying chief features while solving a real world problem but fails to handle uncertainties. Therefore, to analyze real world information system while handling uncertainties, an integrated approach that integrates rough set and cat swarm optimization is proposed. The proposed method identifies the best features concerning degree of dependency of rough set as fitness function in the cat swarm optimization. Further, decision rules are generated using rough set. The feasibility of the suggested research in comparison to existing research is demonstrated by an empirical investigation on the cirrhosis disease. The proposed model that integrates rough set and cat swarm optimization (RSCSO) is analyzed over the detection of cirrhosis disease. It attains an accuracy of 92.06%. Besides, the proposed model RSCSO is compared with the performance of the decision tree and the conventional rough set model. The accuracy attained for decision tree model is 81.6% whereas that of rough set model is 89.08%. Further, the accuracy of RSCSO is compared with cat swarm optimization - decision tree (CSODT) model. The accuracy attained by CSODT model is 86.4%. The analysis shows that the RSCSO model is more accurate as compared to RS, decision tree and CSODT model.