2010年8月30日

SVM - Support Vector Machines

摘要： A Support Vector Machine (SVM) performs classification by constructing an N-dimensional hyperplane that optimally separates the data into two categories. SVM models are closely related to neural networks. In fact, a SVM model using a sigmoid kernel function is equivalent to a two-layer, perceptron neural network. 阅读全文

posted @ 2010-08-30 15:00 leivo 阅读(852) 评论(0) 推荐(0) 编辑

Multilayer Perceptron Neural Networks

摘要： Neural networks are predictive models loosely based on the action of biological neurons. 阅读全文

posted @ 2010-08-30 14:57 leivo 阅读(1077) 评论(0) 推荐(0) 编辑