ItemArticle

Convolutional Neural Network based Di-Strategy Cheetah Optimization Algorithm for Automatic Diabetes Prediction

Title
Convolutional Neural Network based Di-Strategy Cheetah Optimization Algorithm for Automatic Diabetes Prediction
Creator
Kumar, Pankaj; Thomas, Merin; Manur, Manohar; Pani, Alok Kumar
Description
Diabetes is a chronic metabolic disease characterized by elevated blood sugar levels. Diabetes prediction leverages patient data to assess the risk of developing the condition, facilitating early diagnosis and intervention. However, existing models struggle to capture the complex interactions between risk factors due to limited feature representation, leading to inaccurate predictions. This research proposes a Convolutional Neural Network-based Di-Strategy Cheetah Optimization Algorithm (CNN-DS-COA) for automatic diabetes prediction using patient data. The COA is enhanced with tent chaotic mapping and an adaptive search agent, which improves population diversity distribution and convergence speed. Initially, the Pima Indians Diabetes Database (PIMA) and Germany datasets are employed to evaluate the performance of CNN-DS-COA. Min-max normalization is applied to scale the data within a uniform range while preserving relationships among values. The CNN is then used for automatic diabetes prediction, with DS-COA fine-tuning the CNNs parameter values effectively using two strategies. The proposed CNN-DS-COA achieves superior accuracy, with 99.90% and 99.72% on the PIMA and Frankfurt Hospital, Germany datasets, respectively, outperforming existing methods such as stacked ensemble approaches and statistical predictive models. 2025, Research Institute of Intelligent Computer Systems. All rights reserved.
Source
International Journal of Computing;Volume;24;Issue;2;pp.328-335
Date
01-01-2025
Publisher
Research Institute of Intelligent Computer Systems
Subject
adaptive search agent; convolutional neural network; di-strategy cheetah optimization algorithm; min-max normalization; tent chaotic mapping
Coverage
Kumar P., Department of Computer Science and Engineering, Government Engineering College, Vaishali, India; Thomas M., School of Computer science, RV University, Bengaluru, India; Manur M., Department of Computer Science and Engineering, CHRIST (Deemed to be University), Bengaluru, India; Pani A.K., Birla School of Applied Sciences, Birla Global University, Bhubaneswar, India
Rights
All Open Access; Gold Open Access
Relation
ISSN: 17276209;
Format
online
Language
English
Type
Article

Citation

Kumar, Pankaj; Thomas, Merin; Manur, Manohar; Pani, Alok Kumar, “Convolutional Neural Network based Di-Strategy Cheetah Optimization Algorithm for Automatic Diabetes Prediction,” CHRIST (Deemed To Be University) Institutional Repository, accessed June 18, 2026, https://archives.christuniversity.in/items/show/23625.

Output Formats