改进蝙蝠算法优化极限学习机的图像分类

Image classification based on extreme learning machine optimized by improved bat algorithm

针对分类器中的极限学习机参数优化问题,本文提出一种改进蝙蝠算法优化极限学习机的图像分类模型。首先将极限学习机参数看作蝙蝠位置,然后采用改进蝙蝠算法进行求解。采用病毒群体感染主群体,主群体在历代个体间纵向传递信息,病毒群体通过感染操作在同代个体间横向传递信息,增强了算法跳出局部极小值的能力。最后根据最优参数建立图像分类模型,并对模型的性能进行仿真测试。仿真结果表明,相对于对比模型,本文模型不仅提高了图像分类正确率,而且加快了分类速度,是一种有效的图像分类模型。

This paper proposed mage classification model based on the extreme learning machine and bat algorithm.Firstly,the ELM parameters are taken as bat,and then it is solved by the improved bat algorithm which the main groups which consists of bats transmit information cross the vertical generations and the virus groups＇ transfer evolutionary information cross the same generation through virus infection,and the performance of the model is tested.The simulation results show that,compared with the other models,the proposed model not only improves the image classification accuracy,but also accelerate the classification speed,so it is an effective image classification model.