Matlab深度学习实践之手写体识别（含详细注释）

璇焱如柳

已于 2022-03-24 09:52:08 修改

阅读量7.7k

点赞数 18

分类专栏： matlab 深度学习实践文章标签：神经网络深度学习

于 2021-01-07 20:59:21 首次发布

本文链接：https://blog.csdn.net/a1367666195/article/details/112335556

版权

matlab 同时被 2 个专栏收录

15 篇文章 30 订阅

订阅专栏

深度学习实践

3 篇文章 3 订阅

订阅专栏

Matlab这几年在人工智能这块儿也越做越好了，最近为了熟悉matlab如何搭建神经网络，自己做了一个手写体识别实验，记录一下。

实验任务非常简单，网络搭的也非常随意，不合理的地方也懒得改，旨在走通matlab搭建神经网络的流程。

首先，数据集为MNIST数据集
在这里插入图片描述
我已经把数据按类别分好，分为train和test，底下又都有十个子文件夹存放手写体图像。

网络训练代码如下：

clear;close all;clc;
%% 数据读取、增强
   %读取训练集
path_train = 'D:\work\过期文件\手写体识别\MNIST\train';  %训练集路径
folders_train = fullfile(path_train,{'0' '1' '2' '3' '4' '5' '6' '7' '8' '9'}); %读取子目录
imds_train = imageDatastore(folders_train,'FileExtensions','.jpg',...
                            'LabelSource','foldernames');                 %读取所有图像路径
[imdsTrain,imdsValidation] = splitEachLabel(imds_train,0.9,0.1);          %拆分出验证集
   %读取测试集
path_test = 'D:\work\过期文件\手写体识别\MNIST\test';
folders_test = fullfile(path_test,{'0' '1' '2' '3' '4' '5' '6' '7' '8' '9'});
imds_test = imageDatastore(folders_test,'FileExtensions','.jpg',...
                           'LabelSource','foldernames');
   %图像增强
pixelRange = [-2 2];   %平移范围
scaleRange = [0.9 1.1];  %缩放范围
imageAugmenter = imageDataAugmenter( ...
    'RandXTranslation',pixelRange, ...
    'RandYTranslation',pixelRange, ...
    'RandXScale',scaleRange, ...
    'RandYScale',scaleRange);     %定义图像增强器
augimdsTrain = augmentedImageDatastore([28,28],imds_train, ...
    'DataAugmentation',imageAugmenter);  %图像增强
%% 设计（或者读取）网络
layers = [
    imageInputLayer([28 28 1],"Name","imageinput")
    convolution2dLayer([5 5],32,"Name","conv_1","Padding","same","Stride",[2 2])
    reluLayer("Name","relu_1")
    batchNormalizationLayer("Name","batchnorm_1")
    convolution2dLayer([3 3],32,"Name","conv_2","Padding","same")
    reluLayer("Name","relu_2")
    fullyConnectedLayer(512,"Name","fc_1")
    batchNormalizationLayer("Name","batchnorm_2")
    reluLayer("Name","relu_3")
    fullyConnectedLayer(10,"Name","fc_2")
    softmaxLayer("Name","softmax")
    classificationLayer("Name","classoutput")];
% analyzeNetwork(layers)  %分析网络

%% 训练网络
options = trainingOptions('sgdm', ...
    'MiniBatchSize',512, ...
    'MaxEpochs',1, ...
    'InitialLearnRate',1e-2, ...
    'Shuffle','every-epoch', ...
    'ValidationData',imdsValidation, ...
    'ValidationFrequency',3, ...
    'Verbose',1, ...
    'Plots','training-progress');   %设置训练策略
trainedNet = trainNetwork(augimdsTrain,layers,options);  %训练

%% 测试模型
[YPred,probs] = classify(trainedNet,imds_test); 
accuracy = mean(YPred == imds_test.Labels)