一种基于大脑情感学习的快速分类改进算法

为了提高数据分类的快速性与准确性，本文在大脑情感学习（Brain Emotional Learning，BEL）模型的基础上，结合遗传算法（Genetic Algorithm，GA），提出了一种基于GA-BEL的快速分类改进算法.BEL模型根据大脑中杏仁体和眶额皮质之间相互学习的神经生物学原理建立，模拟了情感刺激在大脑短反射通路中被快速处理的过程.因此，基于BEL模型的网络运算速度快.进一步采用遗传算法优化BEL网络权值，提高其分类正确率.在UCI数据集上的对比实验结果表明，无论对于小样本还是大样本数据集，较其他分类算法，GA-BEL算法均有较高的分类正确率和计算效率.

Improved Brain Emotional Learning Algorithm for Fast Classification

An improved fast classification algorithm based on Brain Emotional Learning (BEL) and Genetic Algorithm (GA) is proposed to enhance the accuracy and efficiency.Inspired by the neurobiology research of emotional learning mechanism in amygdala and orbitofrontal cortex,the BEL model is constructed to mimic the mechanism of emotional stimulus processing in human brain.For the short path in the emotional brain,BEL can speed up the learning process.Furthermore,the learning weights in BEL are optimized by GA in order to improve the accuracy.Experiments using UCI datasets are performed,by which the results show that GA-BEL classification obtains higher accuracy and less computing time compared to other classifiers,in both small and large sample datasets.