A multi-objective optimization function is designed to optimize the classification accuracy and the number of selected features jointly, and a single-objective fitness function is constructed by employing the linear weighting method. Then, a Chebyshev opposite-based binary bat algorithm (COBBA) is proposed to mitigate the shortcomings of the bat algorithm (BA) in optimization, such as the lack of population diversity, thereby improving the performance of feature selection. Specifically, a chaotic sequence is generated using a chaotic map to ensure greater diversity in the initial solutions. Additionally, opposite-based learning techniques are used to optimize the initial solutions further and obtain a set of solutions closer to the optimal solution. Chebyshev polynomials are used to improve the quality of the optimal solution by tuning the parameters. Finally, the experiments are conducted on 15 classical datasets. The results show that COBBA outperforms several commonly used swarm intelligence algorithms in average fitness, classification accuracy, and the number of selected features.
(2)提出了一种基于切比雪夫反向的二进制蝙蝠算法(Chebyshev opposite-based binary bat algorithm,COBBA),用于解决所制定的适应度函数。COBBA引入了混沌映射、反向学习技术,并通过切比雪夫多项式调整参数。此外,它还结合了S型函数,将连续解空间转化为离散空间,以适应特征选择任务。
蝙蝠算法由于收敛速度过快,无法全面探索整个搜索空间,探索能力较弱,无法确定全局最优解。Seetharaman等[16]提出贪婪交叉的二进制蝙蝠算法(Binary bat algorith,BBA),解决了获取高分类准确度的简化基因特征的多目标问题。然而,上述工作采用单一的准确性目标衡量性能,这导致了两个不足之处。一方面,特征选择结果倾向于高维解决方案,可能导致过拟合;另一方面,模型的泛化能力较差,所选特征集不稳定。
NaikA K, KuppiliV, EdlaD R. Efficient feature selection using one-pass generalized classifier neural network and binary bat algorithm with a novel fitness function[J]. Soft Computing, 2020, 24: 4575-4587.
[2]
AlazzamH, ShariehA, SabriK E. A feature selection algorithm for intrusion detection system based on pigeon inspired optimizer[J]. Expert Systems with Applications, 2020, 148: 113249.
[3]
CekikR. A new filter feature selection method for text classification[J]. IEEE Access, 2024, 12: 139316-139335.
[4]
ZhangX, ZhangQ, LiP, et al. Multi-feature extraction and selection method to diagnose burn depth from burn images[J]. Electronics, 2024, 13(18): 3665.
[5]
NguyenB H, XueB, ZhangM. A survey on swarm intelligence approaches to feature selection in data mining[J]. Swarm and Evolutionary Computation, 2020, 54: 100663.
[6]
XuX, WuX. Feature selection under orthogonal regression with redundancy minimizing[C]∥IEEE International Conference on Acoustics, Speech and Signal Processing, Barcelona, Spain, 2020: 3457-3461.
[7]
XueY, ZhuH, NeriF. A feature selection approach based on NSGA-II with ReliefF[J]. Applied Soft Computing, 2023, 134: 109987.
[8]
CaoH, YanB, DongL, et al. Multipopulation whale optimization-based feature selection algorithm and its application in human fall detection using inertial measurement unit sensors[J]. Sensors, 2024, 24(24): 24247879.
ZhangMeng-ting, DuJian-qiang, LuoJi-gen, et al. A review of multi-objective optimization in feature selection[J]. Computer Engineering and Applications, 2023, 59(3): 23-32.
[11]
YangX S, GandomiA H. Bat algorithm: a novel approach for global engineering optimization[J]. Engineering Computations, 2012, 29: 464-483.
[12]
GriffithsC A, GiannettiC, AndrzejewskiK T, et al. Comparison of a bat and genetic algorithm generated sequence against lead through programming when assembling a PCB using a six-axis robot with multiple motions and speeds[J]. IEEE Transactions on Industrial Informatics, 2021, 18(2): 1102-1110.
[13]
GhanemW A H M, JantanA. An enhanced bat algorithm with mutation operator for numerical optimization problems[J]. Neural Computing and Applications, 2019, 31: 617-651.
[14]
YangX S. A new metaheuristic bat-inspired algorithm[C]∥Nature Inspired Cooperative Strategies for Optimization, Berlin, Germany, 2010: 65-74.
[15]
SharmaJ, AnnappaB. Community detection using meta-heuristic approach: bat algorithm variants[C]∥Ninth International Conference on Contemporary Computing, Noida, India, 2016: 1-7.
[16]
RajalaxmiR R, RameshA. Binary bat approach for effective spam classification in online social networks[J]. Australian Journal of Basic and Applied Sciences, 2014, 8(18): 383-388.
[17]
SeetharamanA, SundersinghA C. Gene selection and classification using correlation feature selection based binary bat algorithm with greedy crossover[J]. Concurrency and Computation: Practice and Experience, 2022, 34(5): e6718.
[18]
WolpertD H, MacreadyW G. No free lunch theorems for optimization[J]. IEEE Transactions on Evolutionary Computation, 1997, 1(1): 67-82.
[19]
SashiR K, MirunaJ A S A. Efficient feature subset selection and classification using levy flight-based cuckoo search optimization with parallel support vector machine for the breast cancer data[J]. International Journal of Imaging Systems and Technology, 2022, 32(3): 869-881.
[20]
NaghibiT, HoffmannS, PfisterB. Convex approximation of the NP-hard search problem in feature subset selection[C]∥IEEE International Conference on Acoustics, Speech and Signal Processing, Vancouver, Canada, 2013: 3273-3277.
[21]
NguyenB H, XueB, AndreaeP, et al. Multiple reference points-based decomposition for multiobjective feature selection in classification: static and dynamic mechanisms[J]. IEEE Transactions on Evolutionary Computation, 2019, 24(1): 170-184.
[22]
Abd ElazizM, MirjaliliS. A hyper-heuristic for improving the initial population of whale optimization algorithm[J]. Knowledge-Based Systems, 2019, 172: 42-63.
[23]
MasonJ C, HandscombD C. Chebyshev Polynomials[M]. Boca Raton: Chapman and Hall/CRC, 2002.
[24]
MirjaliliS. Dragonfly algorithm: a new meta-heuristic optimization technique for solving single-objective, discrete, and multi-objective problems[J]. Neural Computing and Applications, 2016, 27: 1053-1073.
[25]
KennedyJ, EberhartR. Particle swarm optimization[J]. Proceedings of ICNN'95-International Conference on Neural Networks, 1995, 4: 1942-1948.
[26]
MirjaliliS, MirjaliliS M, LewisA. Grey wolf optimizer[J]. Advances in Engineering Software, 2014, 69: 46-61.