SCALING LAWS IN LEARNING OF CLASSIFICATION TASKS
SCALING LAWS IN LEARNING OF CLASSIFICATION TASKS
复制标题
DOI:
10.1103/physrevlett.70.3167
复制
发表时间:
1993-05-17
影响因子:
8.6
通讯作者:
SOMPOLINSKY, H
中科院分区:
文献类型:
--
作者:
BARKAI, N;SEUNG, HS;SOMPOLINSKY, H
The effect of the structure of the input distribution on the complexity of learning a pattern classification task is investigated. Using statistical mechanics, we study the performance of a winner-take-all machine at learning to classify points generated by a mixture of K Gaussian distributions (''clusters'') in R(N) with intercluster distance u (relative to the cluster width). In the separation limit u >> 1, the number of examples required for learning scales as NKu(-p), where the exponent p is 2 for zero-temperature Gibbs learning and 4 for the Hebb rule.