Hybrid Algorithms of Whale optimization algorithm and k-nearest neighbor to Predict the liver disease

• Published in: EAI endorsed transactions on context-aware systems and applications. Vol. 6; no. 16; p. 156838
• Main Authors: Hajihashemi, Vahid, Hassani, Zeinab, Dehmajnoonie, Iman, Borna, Keivan
• Format: Journal Article
• Language: English
• Published: Ghent European Alliance for Innovation (EAI) 01.03.2019
• Subjects: Algorithms; Diagnosis; Evolutionary algorithm; K-Nearest Neighbor Algorithm; Liver; Liver Disease; Liver diseases; Machine learning; Medical data; Optimization; Optimization algorithms; Pattern recognition; Whale Optimization Algorithm
• ISSN: 2409-0026; 2409-0026
• DOI: 10.4108/eai.13-7-2018.156838

Liver Disease is one of the most common diseases which can be prevented by early diagnosis and up-todate treatment. Advances in machine learning and intelligence techniques have led to the effective diagnosis and prediction of diseases to improve the treatment of patients and reduce the cost of treatment. Whale Optimization Algorithm is a swarm intelligent technique, inspired by the social behavior of whales. One of the effective classification algorithms is K-Nearest Neighbor which is employed for pattern recognition. This paper was designed to investigate the prediction of Liver Disease using a hybrid algorithm including KNN and WOA. In order to evaluate the efficiency of hybrid algorithm, two datasets of liver disease including BUPA and ILPD were used. The results showed that 81.24% and 91.28% of accuracy was gained by the proposed algorithm for BUPA and ILPD, respectively. Experimental results showed that the hybrid WON-KNN is a better classifier to predict the liver diseases.