THE KEY THEOREM OF LEARNING THEORY BASED ON RANDOM SETS SAMPLES

摘要：

Statistical Learning Theory based on random samples is regarded as the best theory for dealing with small-sample learning problems at present And it has become an interesting research after neural networks in machine learning.But it can hardly be handle by the learning problems based on random sets samples.In this paper, combined with the theory of random sets, the definition of the subtraction between the set and the real number is presented, and then some correlative theorems are proven.According to these, some of main concepts of statistical learning theory based on random sets samples are introduced, and at last, the key theorem of learning theory based on random sets samples is given and proven.

关键词： Random sets Hausdorff metric Subtraction ERM principle Key theorem

作者: MING-HU HA LI-FANG ZHENG JI-QIANG CHEN

作者单位: College of Mathematics and Computer Science, Hebei University, Baoding 071002, China

会议类型: 国际会议

会议名称: 2007 International Conference on Machine Learning and Cybernetics(IEEE第六届机器学习与控制论国际会议)

会议地点: 香港

会议语种:英文

页码: 2826-2831

在线出版日期: 2007-08-19（万方平台首次上网日期，不代表论文的发表时间）