GENERALIZATION IN A 2-LAYER NEURAL-NETWORK

被引：27

作者：

KANG, KJ ^{[1
]}

OH, JH ^{[1
]}

KWON, C ^{[1
]}

PARK, Y ^{[1
]}

机构：

[1] MYONG JI UNIV,DEPT PHYS,YONGIN,SOUTH KOREA

来源：

PHYSICAL REVIEW E | 1993年 / 48卷 / 06期

关键词：

D O I：

10.1103/PhysRevE.48.4805

中图分类号：

O35 [流体力学]; O53 [等离子体物理学];

学科分类号：

070204 ; 080103 ; 080704 ;

摘要：

Generalization in a fully connected two-layer neural network with N input nodes, M hidden nodes, a single output node, and binary weights is studied in the annealed approximation. When the number of examples is the order of N, the generalization error approaches a plateau and the system is in a permutation symmetric phase. When the number of examples is of the order of MN, the system undergoes a first-order phase transition to perfect generalisation and the permutation symmetry breaks. Results of the computer simulation show good agreement with analytic calculation

引用

页码：4805 / 4809

页数：5

共 50 条

[31] NEURAL-NETWORK HEURISTICS
JOSIN, G
BYTE, 1987, 12 (11): : 183 - &
[32] A SYNERGETIC NEURAL-NETWORK
NAKAGAWA, M
IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES, 1995, E78A (03) : 412 - 423
[33] Neural-network wavefunctions
McCardle, Kaitlin
NATURE COMPUTATIONAL SCIENCE, 2022, 2 (05): : 284 - 284
[34] Role of stochastic noise and generalization error in the time propagation of neural-network quantum states
Hofmann, Damian
Fabiani, Giammarco
Mentink, Johan H.
Carleo, Giuseppe
Sentef, Michael A.
SCIPOST PHYSICS, 2022, 12 (05):
[35] Forecasting of the chaos by iterations including multi-layer neural-network
Aoyama, T
Zhu, HX
Yoshihara, I
IJCNN 2000: PROCEEDINGS OF THE IEEE-INNS-ENNS INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, VOL IV, 2000, : 467 - 471
[36] Feature Selection Based on a Sparse Neural-Network Layer With Normalizing Constraints
Bugata, Peter
Drotar, Peter
IEEE TRANSACTIONS ON CYBERNETICS, 2023, 53 (01) : 161 - 172
[37] Generalization in a two-layer neural network with multiple outputs
Kang, KJ
Oh, JH
Kwon, C
Park, Y
PHYSICAL REVIEW E, 1996, 54 (02): : 1811 - 1815
[38] On the Effect of Initialization: The Scaling Path of 2-Layer Neural Networks
Neumayer, Sebastian
Chizat, Lenaic
Unser, Michael
JOURNAL OF MACHINE LEARNING RESEARCH, 2024, 25
[39] STORAGE CAPACITY AND LEARNING ALGORITHMS FOR 2-LAYER NEURAL NETWORKS
ENGEL, A
KOHLER, HM
TSCHEPKE, F
VOLLMAYR, H
ZIPPELIUS, A
PHYSICAL REVIEW A, 1992, 45 (10): : 7590 - 7609
[40] Understanding the Generalization Power of Overfitted NTK Models: 3-layer vs. 2-layer
Ju, Peizhong
Lin, Xiaojun
Shroff, Ness B.
2022 58TH ANNUAL ALLERTON CONFERENCE ON COMMUNICATION, CONTROL, AND COMPUTING (ALLERTON), 2022,

← 1 2 3 4 5 →