Local generalization error based monotonic classification extreme learning machine
2017 International Conference on Wavelet Analysis and Pattern Recognition (ICWAPR)(2017)
摘要
The monotonic classification problem is a special case of classification problems, where both the condition attributes and the decision attribute are ordered, and monotonicity relationships existed between them. Based on extreme learning machine, monotonic extreme learning was proposed to solve monotonic classification problems. To improve its generalization capability, in this paper a novel algorithm is proposed based on the local generalization error model, where except for the training error, the objective function takes the sensitivity of the output with respect to the inputs' perturbations into account. An example is conducted to illustrate the feasibility and efficiency of the newly proposed algorithm.
更多查看译文
关键词
Monotonic classification,Generalization capability,Extreme learning machine,Local generalization error,Quadratic programming
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络