{"title":"用于时间序列预测的 K 阶回声型尖峰神经 P 系统","authors":"","doi":"10.1016/j.neucom.2024.128613","DOIUrl":null,"url":null,"abstract":"<div><p>Nonlinear spiking neural P (NSNP) systems are variants of neural-like membrane computing models, abstracted by nonlinear spiking mechanisms of biological neurons. NSNP systems can show rich nonlinear dynamics. This study proposes a new variant of NSNP systems, called <span><math><mi>k</mi></math></span>-order NSNP systems, and derives their mathematical models. The <span><math><mi>k</mi></math></span>-order NSNP systems are able to remember the states of the previous <span><math><mi>k</mi></math></span> moments. Based on the <span><math><mi>k</mi></math></span>-order NSNP systems, we propose a new recurrent-like model, called <span><math><mi>k</mi></math></span>-order echo-type spiking neural P systems or termed kESNP model. Structurally, the <span><math><mi>k</mi></math></span>ESNP model is a <span><math><mi>k</mi></math></span>-order NSNP system equipped with an input layer and an output layer. Inspired by echo state networks (ESN), this <span><math><mi>k</mi></math></span>ESNP model is trained by ridge regression algorithm. Six time series are used as benchmark data sets to evaluate the <span><math><mi>k</mi></math></span>ESNP model and it is compared with 33 baseline prediction methods. The experimental results demonstrate that the proposed <span><math><mi>k</mi></math></span>ESNP model is sufficient for the task of time series forecasting.</p></div>","PeriodicalId":19268,"journal":{"name":"Neurocomputing","volume":null,"pages":null},"PeriodicalIF":5.5000,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"K-order echo-type spiking neural P systems for time series forecasting\",\"authors\":\"\",\"doi\":\"10.1016/j.neucom.2024.128613\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><p>Nonlinear spiking neural P (NSNP) systems are variants of neural-like membrane computing models, abstracted by nonlinear spiking mechanisms of biological neurons. NSNP systems can show rich nonlinear dynamics. This study proposes a new variant of NSNP systems, called <span><math><mi>k</mi></math></span>-order NSNP systems, and derives their mathematical models. The <span><math><mi>k</mi></math></span>-order NSNP systems are able to remember the states of the previous <span><math><mi>k</mi></math></span> moments. Based on the <span><math><mi>k</mi></math></span>-order NSNP systems, we propose a new recurrent-like model, called <span><math><mi>k</mi></math></span>-order echo-type spiking neural P systems or termed kESNP model. Structurally, the <span><math><mi>k</mi></math></span>ESNP model is a <span><math><mi>k</mi></math></span>-order NSNP system equipped with an input layer and an output layer. Inspired by echo state networks (ESN), this <span><math><mi>k</mi></math></span>ESNP model is trained by ridge regression algorithm. Six time series are used as benchmark data sets to evaluate the <span><math><mi>k</mi></math></span>ESNP model and it is compared with 33 baseline prediction methods. The experimental results demonstrate that the proposed <span><math><mi>k</mi></math></span>ESNP model is sufficient for the task of time series forecasting.</p></div>\",\"PeriodicalId\":19268,\"journal\":{\"name\":\"Neurocomputing\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":5.5000,\"publicationDate\":\"2024-09-16\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Neurocomputing\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0925231224013845\",\"RegionNum\":2,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Neurocomputing","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0925231224013845","RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0
摘要
非线性尖峰神经 P(NSNP)系统是类神经膜计算模型的变体,由生物神经元的非线性尖峰机制抽象而来。NSNP 系统可以显示丰富的非线性动态。本研究提出了一种新的 NSNP 系统变体,称为 k 阶 NSNP 系统,并推导出其数学模型。k 阶 NSNP 系统能够记忆前 k 个时刻的状态。在 k 阶 NSNP 系统的基础上,我们提出了一种新的类循环模型,称为 k 阶回声型尖峰神经 P 系统或 kESNP 模型。从结构上讲,kESNP 模型是一个 k 阶 NSNP 系统,配有一个输入层和一个输出层。受回声状态网络(ESN)的启发,该 kESNP 模型采用脊回归算法进行训练。我们使用六个时间序列作为基准数据集来评估 kESNP 模型,并将其与 33 种基准预测方法进行比较。实验结果表明,所提出的 kESNP 模型足以胜任时间序列预测任务。
K-order echo-type spiking neural P systems for time series forecasting
Nonlinear spiking neural P (NSNP) systems are variants of neural-like membrane computing models, abstracted by nonlinear spiking mechanisms of biological neurons. NSNP systems can show rich nonlinear dynamics. This study proposes a new variant of NSNP systems, called -order NSNP systems, and derives their mathematical models. The -order NSNP systems are able to remember the states of the previous moments. Based on the -order NSNP systems, we propose a new recurrent-like model, called -order echo-type spiking neural P systems or termed kESNP model. Structurally, the ESNP model is a -order NSNP system equipped with an input layer and an output layer. Inspired by echo state networks (ESN), this ESNP model is trained by ridge regression algorithm. Six time series are used as benchmark data sets to evaluate the ESNP model and it is compared with 33 baseline prediction methods. The experimental results demonstrate that the proposed ESNP model is sufficient for the task of time series forecasting.
期刊介绍:
Neurocomputing publishes articles describing recent fundamental contributions in the field of neurocomputing. Neurocomputing theory, practice and applications are the essential topics being covered.