Journal of Bionic Engineering ›› 2024, Vol. 21 ›› Issue (1): 409-425.doi: 10.1007/s42235-023-00436-9

• • 上一篇    下一篇

Improved Manta Ray Foraging Optimizer‑based SVM for Feature Selection Problems: A Medical Case Study

Adel Got1,2; Djaafar Zouache2,3; Abdelouahab Moussaoui4; Laith Abualigah5,7,8,9,10,11,12; Ahmed Alsayat6   

  1. 1 Faculty of informatics, University of Science and Technology Houari Boumediene, Algiers, Algeria  2 LRIA Laboratory, University of Science and Technology Houari Boumediene, Algiers, Algeria  3 Computer Science Department, University of Mohamed El Bachir El Ibrahimi, Bordj Bou Arreridj, Algeria  4 Computer Science Department, University of Ferhat Abbas, Setif, Algeria  5 Computer Science Department, Prince Hussein Bin Abdullah Faculty for Information Technology, Al al-Bayt University, Mafraq 25113, Jordan  6 Department of Computer Science, College of Computer and Information Sciences, Jouf University, Jouf, Saudi Arabia  7 Department of Electrical and Computer Engineering, Lebanese American University, 13-5053, Byblos, Lebanon  8 Hourani Center for Applied Scientifc Research, Al-Ahliyya Amman University, Amman 19328, Jordan  9 MEU Research Unit, Middle East University, Amman 11831, Jordan  10 Applied science research center, Applied science private university, Amman 11931, Jordan  11 School of Computer Sciences, Universiti Sains Malaysia, 11800 Penang, Malaysia  12 School of Engineering and Technology, Sunway University Malaysia, 27500 Petaling Jaya, Malaysia
  • 出版日期:2024-01-16 发布日期:2024-02-26
  • 通讯作者: Djaafar Zouache E-mail:djaafarzouache@yahoo.fr
  • 作者简介:Adel Got1,2; Djaafar Zouache2,3; Abdelouahab Moussaoui4; Laith Abualigah5,7,8,9,10,11,12; Ahmed Alsayat6

Improved Manta Ray Foraging Optimizer‑based SVM for Feature Selection Problems: A Medical Case Study

Adel Got1,2; Djaafar Zouache2,3; Abdelouahab Moussaoui4; Laith Abualigah5,7,8,9,10,11,12; Ahmed Alsayat6   

  1. 1 Faculty of informatics, University of Science and Technology Houari Boumediene, Algiers, Algeria  2 LRIA Laboratory, University of Science and Technology Houari Boumediene, Algiers, Algeria  3 Computer Science Department, University of Mohamed El Bachir El Ibrahimi, Bordj Bou Arreridj, Algeria  4 Computer Science Department, University of Ferhat Abbas, Setif, Algeria  5 Computer Science Department, Prince Hussein Bin Abdullah Faculty for Information Technology, Al al-Bayt University, Mafraq 25113, Jordan  6 Department of Computer Science, College of Computer and Information Sciences, Jouf University, Jouf, Saudi Arabia  7 Department of Electrical and Computer Engineering, Lebanese American University, 13-5053, Byblos, Lebanon  8 Hourani Center for Applied Scientifc Research, Al-Ahliyya Amman University, Amman 19328, Jordan  9 MEU Research Unit, Middle East University, Amman 11831, Jordan  10 Applied science research center, Applied science private university, Amman 11931, Jordan  11 School of Computer Sciences, Universiti Sains Malaysia, 11800 Penang, Malaysia  12 School of Engineering and Technology, Sunway University Malaysia, 27500 Petaling Jaya, Malaysia
  • Online:2024-01-16 Published:2024-02-26
  • Contact: Djaafar Zouache E-mail:djaafarzouache@yahoo.fr
  • About author:Adel Got1,2; Djaafar Zouache2,3; Abdelouahab Moussaoui4; Laith Abualigah5,7,8,9,10,11,12; Ahmed Alsayat6

摘要: Support Vector Machine (SVM) has become one of the traditional machine learning algorithms the most used in prediction and classifcation tasks. However, its behavior strongly depends on some parameters, making tuning these parameters a sensitive step to maintain a good performance. On the other hand, and as any other classifer, the performance of SVM is also afected by the input set of features used to build the learning model, which makes the selection of relevant features an important task not only to preserve a good classifcation accuracy but also to reduce the dimensionality of datasets. In this paper, the MRFO + SVM algorithm is introduced by investigating the recent manta ray foraging optimizer to fne-tune the SVM parameters and identify the optimal feature subset simultaneously. The proposed approach is validated and compared with four SVM-based algorithms over eight benchmarking datasets. Additionally, it is applied to a disease Covid-19 dataset. The experimental results show the high ability of the proposed algorithm to fnd the appropriate SVM’s parameters, and its acceptable performance to deal with feature selection problem.

关键词: Support vector machine , · Parameters tuning , · Feature selection , · Bioinspired algorithms , · Manta ray foraging optimizer

Abstract: Support Vector Machine (SVM) has become one of the traditional machine learning algorithms the most used in prediction and classifcation tasks. However, its behavior strongly depends on some parameters, making tuning these parameters a sensitive step to maintain a good performance. On the other hand, and as any other classifer, the performance of SVM is also afected by the input set of features used to build the learning model, which makes the selection of relevant features an important task not only to preserve a good classifcation accuracy but also to reduce the dimensionality of datasets. In this paper, the MRFO + SVM algorithm is introduced by investigating the recent manta ray foraging optimizer to fne-tune the SVM parameters and identify the optimal feature subset simultaneously. The proposed approach is validated and compared with four SVM-based algorithms over eight benchmarking datasets. Additionally, it is applied to a disease Covid-19 dataset. The experimental results show the high ability of the proposed algorithm to fnd the appropriate SVM’s parameters, and its acceptable performance to deal with feature selection problem.

Key words: Support vector machine , · Parameters tuning , · Feature selection , · Bioinspired algorithms , · Manta ray foraging optimizer