A Dendritic Neuron Model Optimized by Meta-Heuristics with a Power-Law-Distributed Population Interaction Network for Financial Time-Series Forecasting
The famous McCulloch–Pitts neuron model has been criticized for being overly simplistic in the long term. At the same time, the dendritic neuron model (DNM) has been shown to be effective in prediction problems, and it accounts for the nonlinear information-processing capacity of synapses and dendri...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-03-01
|
Series: | Mathematics |
Subjects: | |
Online Access: | https://www.mdpi.com/2227-7390/11/5/1251 |
_version_ | 1797614793357524992 |
---|---|
author | Yuxin Zhang Yifei Yang Xiaosi Li Zijing Yuan Yuki Todo Haichuan Yang |
author_facet | Yuxin Zhang Yifei Yang Xiaosi Li Zijing Yuan Yuki Todo Haichuan Yang |
author_sort | Yuxin Zhang |
collection | DOAJ |
description | The famous McCulloch–Pitts neuron model has been criticized for being overly simplistic in the long term. At the same time, the dendritic neuron model (DNM) has been shown to be effective in prediction problems, and it accounts for the nonlinear information-processing capacity of synapses and dendrites. Furthermore, since the classical error back-propagation (BP) algorithm typically experiences problems caused by the overabundance of saddle points and local minima traps, an efficient learning approach for DNMs remains desirable but difficult to implement. In addition to BP, the mainstream DNM-optimization methods include meta-heuristic algorithms (MHAs). However, over the decades, MHAs have developed a large number of different algorithms. How to screen suitable MHAs for optimizing DNMs has become a hot and challenging area of research. In this study, we classify MHAs into different clusters with different population interaction networks (PINs). The performance of DNMs optimized by different clusters of MHAs is tested in the financial time-series-forecasting task. According to the experimental results, the DNM optimized by MHAs with power-law-distributed PINs outperforms the DNM trained based on the BP algorithm. |
first_indexed | 2024-03-11T07:17:15Z |
format | Article |
id | doaj.art-0ff310594aed46038e3d46e9ef0a60f7 |
institution | Directory Open Access Journal |
issn | 2227-7390 |
language | English |
last_indexed | 2024-03-11T07:17:15Z |
publishDate | 2023-03-01 |
publisher | MDPI AG |
record_format | Article |
series | Mathematics |
spelling | doaj.art-0ff310594aed46038e3d46e9ef0a60f72023-11-17T08:10:16ZengMDPI AGMathematics2227-73902023-03-01115125110.3390/math11051251A Dendritic Neuron Model Optimized by Meta-Heuristics with a Power-Law-Distributed Population Interaction Network for Financial Time-Series ForecastingYuxin Zhang0Yifei Yang1Xiaosi Li2Zijing Yuan3Yuki Todo4Haichuan Yang5Department of Engineering, Wesoft Company Ltd., Kawasaki-shi 210-0024, JapanFaculty of Engineering, University of Toyama, Toyama-shi 930-8555, JapanDepartment of Engineering, Wesoft Company Ltd., Kawasaki-shi 210-0024, JapanFaculty of Engineering, University of Toyama, Toyama-shi 930-8555, JapanFaculty of Electrical and Computer Engineering, Kanazawa University, Kanazawa-shi 920-1192, JapanDepartment of Engineering, Wesoft Company Ltd., Kawasaki-shi 210-0024, JapanThe famous McCulloch–Pitts neuron model has been criticized for being overly simplistic in the long term. At the same time, the dendritic neuron model (DNM) has been shown to be effective in prediction problems, and it accounts for the nonlinear information-processing capacity of synapses and dendrites. Furthermore, since the classical error back-propagation (BP) algorithm typically experiences problems caused by the overabundance of saddle points and local minima traps, an efficient learning approach for DNMs remains desirable but difficult to implement. In addition to BP, the mainstream DNM-optimization methods include meta-heuristic algorithms (MHAs). However, over the decades, MHAs have developed a large number of different algorithms. How to screen suitable MHAs for optimizing DNMs has become a hot and challenging area of research. In this study, we classify MHAs into different clusters with different population interaction networks (PINs). The performance of DNMs optimized by different clusters of MHAs is tested in the financial time-series-forecasting task. According to the experimental results, the DNM optimized by MHAs with power-law-distributed PINs outperforms the DNM trained based on the BP algorithm.https://www.mdpi.com/2227-7390/11/5/1251dendritic neuron modelmeta-heuristic algorithmsfinancial time-series forecastingpopulation interaction networks |
spellingShingle | Yuxin Zhang Yifei Yang Xiaosi Li Zijing Yuan Yuki Todo Haichuan Yang A Dendritic Neuron Model Optimized by Meta-Heuristics with a Power-Law-Distributed Population Interaction Network for Financial Time-Series Forecasting Mathematics dendritic neuron model meta-heuristic algorithms financial time-series forecasting population interaction networks |
title | A Dendritic Neuron Model Optimized by Meta-Heuristics with a Power-Law-Distributed Population Interaction Network for Financial Time-Series Forecasting |
title_full | A Dendritic Neuron Model Optimized by Meta-Heuristics with a Power-Law-Distributed Population Interaction Network for Financial Time-Series Forecasting |
title_fullStr | A Dendritic Neuron Model Optimized by Meta-Heuristics with a Power-Law-Distributed Population Interaction Network for Financial Time-Series Forecasting |
title_full_unstemmed | A Dendritic Neuron Model Optimized by Meta-Heuristics with a Power-Law-Distributed Population Interaction Network for Financial Time-Series Forecasting |
title_short | A Dendritic Neuron Model Optimized by Meta-Heuristics with a Power-Law-Distributed Population Interaction Network for Financial Time-Series Forecasting |
title_sort | dendritic neuron model optimized by meta heuristics with a power law distributed population interaction network for financial time series forecasting |
topic | dendritic neuron model meta-heuristic algorithms financial time-series forecasting population interaction networks |
url | https://www.mdpi.com/2227-7390/11/5/1251 |
work_keys_str_mv | AT yuxinzhang adendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT yifeiyang adendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT xiaosili adendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT zijingyuan adendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT yukitodo adendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT haichuanyang adendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT yuxinzhang dendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT yifeiyang dendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT xiaosili dendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT zijingyuan dendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT yukitodo dendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting AT haichuanyang dendriticneuronmodeloptimizedbymetaheuristicswithapowerlawdistributedpopulationinteractionnetworkforfinancialtimeseriesforecasting |