Formation of local heterogeneity under energy collection in neural networks

被引:0
作者
Ying Xie
Zhao Yao
Jun Ma
机构
[1] Lanzhou University of Technology,Department of Physics
[2] Central South University,School of Physics and Electronics
[3] Chongqing University of Posts and Telecommunications,School of Science
来源
Science China Technological Sciences | 2023年 / 66卷
关键词
energy balance; synapse coupling; heterogeneity; neural network;
D O I
暂无
中图分类号
学科分类号
摘要
Static distribution of intracellular and extracellular ions can induce the spatial distribution of electric field in the cell, while stochastic diffusion and propagation of ions can generate channel current accompanied by the magnetic field. In fact, the field energy (including magnetic and electric fields) in each cell or neuron can be changed under external stimuli or the deformation of shape. Furthermore, energy pumping occurs, and the synaptic connection is created adaptively to keep energy balance when more neurons are clustered in the same region because of electromagnetic field superposition. For a synchronous and homogeneous network, all neurons keep energy balance with the adjacent neurons, and the network becomes uniform and identical. For cardiac tissue and biological media, local heterogeneity such as sinoatrial node can emit continuous wave front, and target waves are propagated to regulate the neural activities and heartbeat rhythm. In this paper, the Hamilton energy in a simple neuron model is calculated according to the Helmholtz theorem, and more neurons are clustered to build a regular network (chain network and square array). The creation and growth of synapse connection are controlled to pump energy, and local heterogeneity is formed by taming one intrinsic parameter when more energy is accumulated in a few neurons in the network. These results indicate that asymmetrical energy transport will induce the occurrence of heterogeneity in the network, and adaptive synaptic regulation on neurons is activated to keep local energy balance before reaching perfect synchronization. When external stimuli with diversity in intensity are applied, neurons are excited with energy diversity, and possible shape deformation is induced to keep local energy balance. As a result, heterogeneity is developed in a local area.
引用
收藏
页码:439 / 455
页数:16
相关论文
共 50 条
[41]   Exploiting Artificial Neural Networks for the Prediction of Ancillary Energy Market Prices [J].
Giovanelli, Christian ;
Sierla, Seppo ;
Ichise, Ryutaro ;
Vyatkin, Valeriy .
ENERGIES, 2018, 11 (07)
[42]   Synchronization in fractional-order neural networks by the energy balance strategy [J].
Yao, Zhao ;
Sun, Kehui ;
He, Shaobo .
COGNITIVE NEURODYNAMICS, 2024, 18 (02) :701-713
[43]   Integrated solutions based on neural networks for optimizing energy management in a microgrid [J].
Otilia, Dragomir ;
Florin, Dragomir .
2013 4TH INTERNATIONAL SYMPOSIUM ON ELECTRICAL AND ELECTRONICS ENGINEERING (ISEEE), 2013,
[44]   Interpretation of Mössbauer spectra in the energy and time domain with neural networks [J].
H. Paulsen ;
R. Linder ;
F. Wagner ;
H. Winkler ;
S.J. Pöppl ;
A.X. Trautwein .
Hyperfine Interactions, 2000, 126 :421-424
[45]   Direct prediction of profiles of sequences compatible with a protein structure by neural networks with fragment-based local and energy-based nonlocal profiles [J].
Li, Zhixiu ;
Yang, Yuedong ;
Faraggi, Eshel ;
Zhan, Jian ;
Zhou, Yaoqi .
PROTEINS-STRUCTURE FUNCTION AND BIOINFORMATICS, 2014, 82 (10) :2565-2573
[46]   Temperature prediction of yam under infrared drying based on neural networks [J].
Zhang, Lili ;
Wang, Xiangyou ;
Zhang, Haipeng .
Nongye Jixie Xuebao/Transactions of the Chinese Society for Agricultural Machinery, 2014, 45 (11) :246-249and336
[47]   Distributed Coordinated Attitude Tracking Control for Spacecraft Formation Based on Neural Networks [J].
Wang, Wenjia ;
Li, Chuanjiang ;
Sun, Yanchao ;
Jiang, Boyan .
2017 29TH CHINESE CONTROL AND DECISION CONFERENCE (CCDC), 2017, :7190-7195
[48]   Formation of synchronous activity through STDP in recurrent neural networks with heterogenous delays [J].
Frédéric Henry ;
Emmanuel Daucél .
BMC Neuroscience, 10 (Suppl 1)
[49]   Integration of genetic algorithms and neural networks for the formation of the classifier of the hierarchical Choquet integral [J].
Chen, Chin-Yi ;
Huang, Jih-Jeng .
INFORMATION SCIENCES, 2020, 537 :46-61
[50]   Adaptive Local Update and Neural Composition for Accelerating Federated Learning in Heterogeneous Edge Networks [J].
Liu, Jianchun ;
Yan, Jiaming ;
Qi, Ji ;
Xu, Hongli ;
Wang, Shilong ;
Qiao, Chunming ;
Huang, Liusheng .
IEEE TRANSACTIONS ON NETWORKING, 2025,