Gene Selection Based Cancer Classification With Adaptive Optimization Using Deep Learning Architecture
Early cancer identification using gene expression data is critical for providing successful patient care. Accurate data recognition is essential to prevent improper detection because it may result in higher complexities and increased mortality rates. Gene expression data typically include numerous f...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10507789/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832583958073704448 |
|---|---|
| author | Anju Das N. Neelima K. Deepa Tolga Ozer |
| author_facet | Anju Das N. Neelima K. Deepa Tolga Ozer |
| author_sort | Anju Das |
| collection | DOAJ |
| description | Early cancer identification using gene expression data is critical for providing successful patient care. Accurate data recognition is essential to prevent improper detection because it may result in higher complexities and increased mortality rates. Gene expression data typically include numerous features, each representing distinct genes. The abundance of features introduces high dimensionality, contributing to high computational complexity and resource demands. Furthermore, the presence of redundant or highly correlated selected features may lead to multicollinearity issues. In the existing works, certain limitations, such as reduced performance due to degraded data quality, high storage space requirements, overfitting issues, and lack of robustness, can compromise overall classification accuracy. To address these challenges and enhance classification outcomes, this research employs an efficient framework based on a deep learning (DL) approach. Initially, the data is collected from five gene cancer datasets, which are then augmented to maximize the data size. Min-Max Normalization is used for data pre-processing. The Enhanced Chimp Optimization (ECO) algorithm is applied to select the most significant genes while eliminating redundant or unwanted ones. Based on the selected gene set, the Depth-wise Separable Convolutional Neural Network (DSCNN) is employed to categorize diverse cancerous and non-cancerous classes. The performance of the proposed model is greatly improved by resolving the dimensionality and overfitting problems. The implementation is carried out using PYTHON, and the overall accuracy of the DSCNN model is determined to be 99.18% across all datasets. Additionally, metrics such as precision, recall, and F1 score are evaluated to analyze the model’s overall performance. The proposed model demonstrates a significantly higher effective performance compared to existing approaches. |
| format | Article |
| id | doaj-art-09693f668fbd496097e22806b6317795 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-09693f668fbd496097e22806b63177952025-01-28T00:00:59ZengIEEEIEEE Access2169-35362024-01-0112622346225510.1109/ACCESS.2024.339263310507789Gene Selection Based Cancer Classification With Adaptive Optimization Using Deep Learning ArchitectureAnju Das0https://orcid.org/0009-0002-9858-3840N. Neelima1K. Deepa2Tolga Ozer3Department of Electronics and Communication Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Bengaluru, IndiaDepartment of Electronics and Communication Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Bengaluru, IndiaDepartment of Electrical and Electronics Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Bengaluru, IndiaDepartment of Electric Electronic Engineering, Technology Faculty, Afyon Kocatepe University, Afyonkarahisar, TurkeyEarly cancer identification using gene expression data is critical for providing successful patient care. Accurate data recognition is essential to prevent improper detection because it may result in higher complexities and increased mortality rates. Gene expression data typically include numerous features, each representing distinct genes. The abundance of features introduces high dimensionality, contributing to high computational complexity and resource demands. Furthermore, the presence of redundant or highly correlated selected features may lead to multicollinearity issues. In the existing works, certain limitations, such as reduced performance due to degraded data quality, high storage space requirements, overfitting issues, and lack of robustness, can compromise overall classification accuracy. To address these challenges and enhance classification outcomes, this research employs an efficient framework based on a deep learning (DL) approach. Initially, the data is collected from five gene cancer datasets, which are then augmented to maximize the data size. Min-Max Normalization is used for data pre-processing. The Enhanced Chimp Optimization (ECO) algorithm is applied to select the most significant genes while eliminating redundant or unwanted ones. Based on the selected gene set, the Depth-wise Separable Convolutional Neural Network (DSCNN) is employed to categorize diverse cancerous and non-cancerous classes. The performance of the proposed model is greatly improved by resolving the dimensionality and overfitting problems. The implementation is carried out using PYTHON, and the overall accuracy of the DSCNN model is determined to be 99.18% across all datasets. Additionally, metrics such as precision, recall, and F1 score are evaluated to analyze the model’s overall performance. The proposed model demonstrates a significantly higher effective performance compared to existing approaches.https://ieeexplore.ieee.org/document/10507789/Gene expression datamin-max normalizationenhanced chimp optimizationdepth-wise separable convolutionoptimal genescancer classification |
| spellingShingle | Anju Das N. Neelima K. Deepa Tolga Ozer Gene Selection Based Cancer Classification With Adaptive Optimization Using Deep Learning Architecture IEEE Access Gene expression data min-max normalization enhanced chimp optimization depth-wise separable convolution optimal genes cancer classification |
| title | Gene Selection Based Cancer Classification With Adaptive Optimization Using Deep Learning Architecture |
| title_full | Gene Selection Based Cancer Classification With Adaptive Optimization Using Deep Learning Architecture |
| title_fullStr | Gene Selection Based Cancer Classification With Adaptive Optimization Using Deep Learning Architecture |
| title_full_unstemmed | Gene Selection Based Cancer Classification With Adaptive Optimization Using Deep Learning Architecture |
| title_short | Gene Selection Based Cancer Classification With Adaptive Optimization Using Deep Learning Architecture |
| title_sort | gene selection based cancer classification with adaptive optimization using deep learning architecture |
| topic | Gene expression data min-max normalization enhanced chimp optimization depth-wise separable convolution optimal genes cancer classification |
| url | https://ieeexplore.ieee.org/document/10507789/ |
| work_keys_str_mv | AT anjudas geneselectionbasedcancerclassificationwithadaptiveoptimizationusingdeeplearningarchitecture AT nneelima geneselectionbasedcancerclassificationwithadaptiveoptimizationusingdeeplearningarchitecture AT kdeepa geneselectionbasedcancerclassificationwithadaptiveoptimizationusingdeeplearningarchitecture AT tolgaozer geneselectionbasedcancerclassificationwithadaptiveoptimizationusingdeeplearningarchitecture |