A global neural network learning machine: Coupled integer and fractional calculus operator with an adaptive learning scheme. (November 2021)

• Record Type: Journal Article
• Authors: Zhang, Huaqing; Pu, Yi-Fei; Xie, Xuetao; Zhang, Bingran; Wang, Jian; Huang, Tingwen
• Is Part Of: Neural networks. Volume 143(2021)
• Page Start: 386
• Page End: 399
• Publication Date: 2021-11
• Subjects: Swarm intelligent -- Fractional calculus -- Global optimization -- Neural network -- Adaptive learning rate; Neural computers -- Periodicals; Neural networks (Computer science) -- Periodicals; Neural networks (Neurobiology) -- Periodicals; Nervous System -- Periodicals; Ordinateurs neuronaux -- Périodiques; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux neuronaux (Neurobiologie) -- Périodiques; Neural computers; Neural networks (Computer science); Neural networks (Neurobiology); Periodicals; 006.32
• DOI: 10.1016/j.neunet.2021.06.021
• Languages: English
• ISSNs: 0893-6080

Abstract

Abstract: Find the global optimal solution of the model is one promising research topic in computational intelligent community. Dependent on analogies to natural processes, the evolutionary swarm intelligent algorithms are widely used for solving global optimization problems which directed by the fitness values. In this paper, we propose one efficient fractional global learning machine (Fragmachine) which includes two stages (descending and ascending) to determine the optimal search path. The neural network model is used to approach the given fitness value. Specifically, for the descending stage, the integer gradient of the network output with respect the current location is employed to find the next descending point, while for the ascending stage, the fractional gradient is implemented to climb and escape from the local optimal point. We further propose one adaptive learning rate during training which relies on both the current gradient (integer or fractional) information and the fitness value. Finally, a series of numerical experiments verify the effectiveness of the proposed algorithm, Fragmachine.