基于遗传算法优化BP神经网络的手写数字识别  

Handwritten Digit Recognition Based on Genetic Algorithm Optimizing BP Neural Network

作　　者：张茜 ZHANG Xi(School of Mathematical Sciences,South China Normal University,Guangzhou Guangdong 510631,China)

机构地区：[1]华南师范大学数学科学学院,广东广州510631

出　　处：《信息与电脑》2023年第7期75-77,共3页Information & Computer

摘　　要：手写数字识别是经典的分类任务,在支票阅读、街道编号识别等方面具有许多实际应用。为了提高手写数字分类准确性,文章提出了基于遗传算法(Genetic Algorithm,GA)优化的反向传播(Back Propagation,BP)神经网络模型,即GA-BP神经网络模型。基于MNIST手写数字训练集,GA-BP神经网络模型在迭代50次时能达到95.07%的分类准确率,显著高于BP神经网络等单一分类模型的准确率,验证了改进后的模型在手写数字分类上的有效性。Handwritten digit recognition is a classical classification task with many practical applications in check reading,street number recognition,etc.In order to improve the classification accuracy of handwritten digits,the paper proposes a Back Propagation(BP) neural network model based on Genetic Algorithm(GA) optimization,namely GA-BP neural network model.Based on the MNIST handwritten digit training set,the GA-BP neural network model can achieve 95.07% classification accuracy at 50 iterations,which is significantly higher than the accuracy of a single classification model such as BP neural network,verifying the effectiveness of the improved model for handwritten digit classification.

关 键 词：手写数字识别 反向传播(BP)神经网络 遗传算法(GA) 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]