{"title":"Hopfield网络吸引子的标度定律","authors":"G. Weisbuch, F. Fogelman-Soulié","doi":"10.1051/JPHYSLET:019850046014062300","DOIUrl":null,"url":null,"abstract":"Networks of threshold automata are random dynamical systems with a large number of attractors, which J. Hopfield proposed to use as associative memories. We establish the scaling laws relating the maximum number of «useful» attractors and the radius of the attraction basin to the number of automata. A by-product of our analysis is a better choice for thresholds which doubles the performances in terms of the maximum number of «useful» attractors Les reseaux d'automates a seuil sont des systemes dynamiques a structure aleatoire semblables aux verres de spins dont J. Hopfield a propose l'application comme memoires associatives. Nous etablissons les lois d'echelles reliant le nombre maximum d'attracteurs utiles et la distance d'attraction, au nombre des automates du reseau. Notre approche permet aussi un meilleur choix des seuils, ce qui double les performances du reseau en nombre d'attracteurs.","PeriodicalId":14822,"journal":{"name":"Journal De Physique Lettres","volume":"6 1","pages":"623-630"},"PeriodicalIF":0.0000,"publicationDate":"1985-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"100","resultStr":"{\"title\":\"Scaling laws for the attractors of Hopfield networks\",\"authors\":\"G. Weisbuch, F. Fogelman-Soulié\",\"doi\":\"10.1051/JPHYSLET:019850046014062300\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Networks of threshold automata are random dynamical systems with a large number of attractors, which J. Hopfield proposed to use as associative memories. We establish the scaling laws relating the maximum number of «useful» attractors and the radius of the attraction basin to the number of automata. A by-product of our analysis is a better choice for thresholds which doubles the performances in terms of the maximum number of «useful» attractors Les reseaux d'automates a seuil sont des systemes dynamiques a structure aleatoire semblables aux verres de spins dont J. Hopfield a propose l'application comme memoires associatives. Nous etablissons les lois d'echelles reliant le nombre maximum d'attracteurs utiles et la distance d'attraction, au nombre des automates du reseau. Notre approche permet aussi un meilleur choix des seuils, ce qui double les performances du reseau en nombre d'attracteurs.\",\"PeriodicalId\":14822,\"journal\":{\"name\":\"Journal De Physique Lettres\",\"volume\":\"6 1\",\"pages\":\"623-630\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"1985-07-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"100\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal De Physique Lettres\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1051/JPHYSLET:019850046014062300\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal De Physique Lettres","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1051/JPHYSLET:019850046014062300","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 100
摘要
阈值自动机网络是具有大量吸引子的随机动态系统,Hopfield提出将其作为联想记忆。我们建立了“有用”吸引子的最大数目和吸引盆半径与自动机数目之间的标度律。我们分析的副产品是一个更好的阈值选择,就“有用”吸引子的最大数量而言,阈值的性能增加了一倍。Les reseaux d’automates、seseil sonds、systemes dynamques、结构可变的可组装物、verres de spins和J. Hopfield提出了l’application comcommemorires associative。Nous的建立,如:loes d' echeles,依赖于nombre,最大值d' attracters,效用,距离d'attraction,和nombre des automates du rerees。巴黎的方法允许我们选择美丽的风景,而不是双色的风景,或者是双色的风景。
Scaling laws for the attractors of Hopfield networks
Networks of threshold automata are random dynamical systems with a large number of attractors, which J. Hopfield proposed to use as associative memories. We establish the scaling laws relating the maximum number of «useful» attractors and the radius of the attraction basin to the number of automata. A by-product of our analysis is a better choice for thresholds which doubles the performances in terms of the maximum number of «useful» attractors Les reseaux d'automates a seuil sont des systemes dynamiques a structure aleatoire semblables aux verres de spins dont J. Hopfield a propose l'application comme memoires associatives. Nous etablissons les lois d'echelles reliant le nombre maximum d'attracteurs utiles et la distance d'attraction, au nombre des automates du reseau. Notre approche permet aussi un meilleur choix des seuils, ce qui double les performances du reseau en nombre d'attracteurs.