作者:ZhiweiXu
Z. Liang, W. Liang, Z. Wang, X. Ma, L. Liu and Z. Zhu, "Multiobjective Evolutionary Multitasking With Two-Stage Adaptive Knowledge Transfer Based on Population Distribution," in IEEE Transactions on Systems, Man, and Cybernetics: Systems, doi: 10.1109/TSMC.2021.3096220.
的论文学习笔记,只供学习使用,不作商业用途,侵权删除。并且本人学术功底有限如果有思路不正确的地方欢迎批评指正![21] A. Gupta, Y .-S. Ong, and L. Feng, “Multifactorial evolution: Toward evolutionary multitasking,” IEEE Trans. Evol. Comput., vol. 20, no. 3, pp. 343–357, Jun. 2016.
[22] K. C. Tan, L. Feng, and M. Jiang, “Evolutionary transfer optimization— A new frontier in evolutionary computation research,” IEEE Comput. Intell. Mag., vol. 16, no. 1, pp. 22–33, Feb. 2021.
[23] A. Gupta, Y .-S. Ong, L. Feng, and K. C. Tan, “Multiobjective multifactorial optimization in evolutionary multitasking,” IEEE Trans. Cybern., vol. 47, no. 7, pp. 1652–1665, Jul. 2017.
[24] C. E. Y ang, J. L. Ding, K. C. Tan, and Y . C. Jin, “Two-stage assortative mating for multi-objective multifactorial evolutionary optimization,” in Proc. IEEE 56th Annu. Conf. Decis. Control (CDC), Melbourne, VIC, Australia, Dec. 2017, pp. 76–81.
[25] L. Feng et al., “Evolutionary multitasking via explicit autoencoding,” IEEE Trans. Cybern., vol. 49, no. 9, pp. 3457–3470, Sep. 2019.
[26] Y . L. Chen, J. H. Zhong, and M. K. Tan, “A fast memetic multiobjective differential evolution for multi-tasking optimization,” in Proc. IEEE Congr . Evol. Comput. (CEC), Rio de Janeiro, Brazil, Jul. 2018, pp. 1–8.
[27] N. Q. Tuan, T. D. Hoang, and H. T. T. Binh, “A guided differential evolutionary multi-tasking with powell search method for solving multiobjective continuous optimization,” in Proc. IEEE Congr . Evol. Comput. (CEC), Rio de Janeiro, Brazil, Jul. 2018, pp. 1–8.
[28] Y . Y uan, Y .-S. Ong, A. Gupta, P . S. Tan, and H. Xu, “Evolutionary multitasking in permutation-based combinatorial optimization problems: Realization with TSP , QAP , LOP , and JSP ,” in Proc. IEEE Region 10 Annu. Int. Conf. (TENCON), Singapore, Nov. 2016, pp. 3157–3164.
[29] R. Sagarna and Y .-S. Ong, “Concurrently searching branches in software tests generation through multitask evolution,” in Proc. IEEE Symp. Series Comput. Intell. (SSCI), Athens, Greece, Dec. 2016, pp. 1–8.
[30] R. Chandra, A. Gupta, Y .-S. Ong, and C.-K. Goh, “Evolutionary multitask learning for modular knowledge representation in neural networks,” Neural Process. Lett., vol. 47, no. 3, pp. 993–1009, Jun. 2018.
[31] J. Zhong, L. Feng, W. Cai, and Y .-S. Ong, “Multifactorial genetic programming for symbolic regression problems,” IEEE Trans. Syst., Man, Cybern., Syst., vol. 50, no. 11, pp. 4492–4505, Nov. 2020.
[32] A. Rauniyar, R. Nath, and P . K. Muhuri, “Multi-factorial evolutionary algorithm based novel solution approach for multi-objective pollutionrouting problem,” Comput. Ind. Eng., vol. 130, no. 5, pp. 757–771, Apr. 2019.
[33] H. Li, Y .-S. Ong, M. G. Gong, and Z. K. Wang, “Evolutionary multitasking sparse reconstruction: Framework and case study,” IEEE Trans. Evol. Comput., vol. 23, no. 5, pp. 733–747, Oct. 2019.
【目前存在的问题】EMT算法的研究已经取得了显著的进展,但仍有进一步改进的空间,一些悬而未决的问题仍有待解决。特别是在高度相似的问题上,现有算法可能无法充分利用高质量解的知识来提高种群的收敛性,或者没有很好地处理种群陷入局部最优的情况。在低相似性问题上,任务的总体分布通常是不同的,这往往会导致任务之间的负迁移[34]。
【本文创新点】为了解决上述问题,本文提出了一种新的基于群体分布的两阶段自适应知识转移模型。EMT-PD首先为每个任务建立概率模型,然后从不同概率模型的乘积中获取知识。这些知识有助于加快种群的收敛速度。在知识转移的第一阶段,通过自适应权重调整每个个体的搜索步长,可以降低产生负转移的概率。在知识转移的第二阶段,动态调整每个个体的搜索范围,以促进种群多样性,避免陷入局部最优。EMT-PD然后在多任务多目标优化测试套件上进行测试。与其他最先进的进化多/单任务算法的比较研究证明了EMT-PD的竞争力。本文的主要贡献如下。
1)为了提高优化的效率和性能,提出了一种多任务多目标进化优化算法,该算法采用了一种新的知识提取和传递方法。
2)提出了一个基于著名的MaF【35】的多任务多目标优化测试套件。
[35] M. Q. Li et al., “A benchmark test suite for evolutionary many-objective optimization,” Complex Intell. Syst., vol. 3, no. 1, pp. 67–81, Mar. 2017.
3)基于三个测试套件,通过与其他最先进的算法进行比较,充分分析了EMT-PD 的优缺点。
[21] MFEA: A. Gupta, Y .-S. Ong, and L. Feng, “Multifactorial evolution: Toward evolutionary multitasking,” IEEE Trans. Evol. Comput., vol. 20, no. 3, pp. 343–357, Jun. 2016.
[23] MOMFEA: A. Gupta, Y .-S. Ong, L. Feng, and K. C. Tan, “Multiobjective multifactorial optimization in evolutionary multitasking,” IEEE Trans. Cybern., vol. 47, no. 7, pp. 1652–1665, Jul. 2017.
[24] TMOMFEA: C. E. Y ang, J. L. Ding, K. C. Tan, and Y . C. Jin, “Two-stage assortative mating for multi-objective multifactorial evolutionary optimization,” in Proc. IEEE 56th Annu. Conf. Decis. Control (CDC), Melbourne, VIC, Australia, Dec. 2017, pp. 76–81.
[31] MFGP: J. Zhong, L. Feng, W. Cai, and Y .-S. Ong, “Multifactorial genetic programming for symbolic regression problems,” IEEE Trans. Syst., Man, Cybern., Syst., vol. 50, no. 11, pp. 4492–4505, Nov. 2020.
[36] M-BLEA: A. Gupta, J. Ma´ ndziuk, and Y .-S. Ong, “Evolutionary multitasking in bi-level optimization,” Complex Intell. Syst., vol. 1, no. 1, pp. 83–95, Dec. 2015.
[37] LDA-MFEA: K. K. Bali, A. Gupta, L. Feng, Y . S. Ong, and P . S. Tan, “Linearized domain adaptation in evolutionary multitasking,” in Proc. IEEE Congr . Evol. Comput. (CEC), Donostia, Spain, Jun. 2017, pp. 1295–1302.
[38] S&M-MFEA: B. S. Da, A. Gupta, Y .-S. Ong, and L. Feng, “Evolutionary multitasking across single and multi-objective formulations for improved problem solving,” in Proc. IEEE Congr . Evol. Comput. (CEC), V ancouver, BC, Canada, Jul. 2016, pp. 1695–1701.
[39] GMFEA: J. L. Ding, C. Y ang, Y . C. Jin, and T. Y . Chai, “Generalized multitasking for evolutionary optimization of expensive problems,” IEEE Trans. Evol. Comput., vol. 23, no. 1, pp. 44–58, Feb. 2019.
[40] MTO-DRA: M. G. Gong, Z. D. Tang, H. Li, and J. Zhang, “Evolutionary multitasking with dynamic resource allocating strategy,” IEEE Trans. Evol. Comput., vol. 23, no. 5, pp. 858–869, Oct. 2019.
[41] MFEAII: K. K. Bali, Y .-S. Ong, A. Gupta, and P . S. Tan, “Multifactorial evolutionary algorithm with online transfer parameter estimation: MFEA-II,” IEEE Trans. Evol. Comput., vol. 24, no. 1, pp. 69–83, Feb. 2020.
[42] MFEA-GHS: Z. P . Liang, J. Zhang, L. Feng, and Z. X. Zhu, “A hybrid of genetic transform and hyper-rectangle search strategies for evolutionary multi-tasking,” Expert Syst. Appl., vol. 138, no. 30, Dec. 2019, Art. no. 112798.
[43] B. S. Da, A. Gupta, Y . S. Ong, and L. Feng, The Boon of Gene-Culture Interaction for Effective Evolutionary Multitasking (Lecture Notes in Computer Science, 9592). Cham, Switzerland: Springer, Feb. 2016, pp. 54–65
[46] L. Feng et al., “An empirical study of multifactorial PSO and multifactorial DE,” in Proc. IEEE Congr . Evol. Comput. (CEC), Donostia, Spain, Jun. 2017, pp. 921–928.
[47] Z. D. Tang and M. G. Gong, “Adaptive multifactorial particle swarm optimisation,” CAAI Trans. Intell. Technol., vol. 4, no. 1, pp. 37–46, Mar. 2019.
[48] H. Song, A. K. Qin, P .-W. Tsai, and J. J. Liang, “Multitasking multiswarm optimization,” in Proc. IEEE Congr . Evol. Comput. (CEC), Wellington, New Zealand, Jun. 2019, pp. 1937–1944.
[51] D. N. Liu, S. J. Huang, and J. H. Zhong, “Surrogate-assisted multitasking memetic algorithm,” in Proc. IEEE Congr . Evol. Comput. (CEC), Rio de Janeiro, Brazil, Jul. 2018, pp. 1–8.
[52] MFDE: L. Zhou et al., “Towards effective mutation for knowledge transfer in multifactorial differential evolution,” in Proc. IEEE Congr . Evol. Comput. (CEC), Wellington, New Zealand, Jun. 2019, pp. 1541–1547.
[53] MFEA-SADE: Z. Liang, H. Dong, C. Liu, W. Liang, and Z. Zhu, “Evolutionary multitasking for multiobjective optimization with subspace alignment and adaptive differential evolution,” IEEE Trans. Cybern., early access, Jun. 24, 2020, doi: 10.1109/TCYB.2020.2980888.
[25] L. Feng et al., “Evolutionary multitasking via explicit autoencoding,” IEEE Trans. Cybern., vol. 49, no. 9, pp. 3457–3470, Sep. 2019.
【ML and EC】详解EM算法与混合高斯模型Gaussian mixture model, GMM
【ML and EC】二维高斯分布Two-dimensional Gaussian distribution的参数分析
At the first stage, the search step size of each individual is adjusted adaptively to reduce the impact of negative transfer. At the second stage, the search range of each individual is further adjusted based on an intermediate individual, which increases the population diversity of the population and avoids getting trapped in local optimum.