資源簡介
利用基于tensorflow2的keras框架,搭建CNN卷積神經(jīng)網(wǎng)絡(luò)模型,對(duì)手寫數(shù)字識(shí)別數(shù)據(jù)集mnist進(jìn)行分類,網(wǎng)絡(luò)規(guī)模小,訓(xùn)練精度高。網(wǎng)絡(luò)包括三個(gè)卷積層,兩個(gè)池化層和全連接層,在測試集上實(shí)現(xiàn)了99%左右的識(shí)別率。
代碼片段和文件信息
from?keras.datasets?import?mnist
from?keras.layers?import?Dense?Conv2D?MaxPooling2D?Flatten
from?keras.models?import?Sequential
from?keras.utils?import?to_categorical
“““數(shù)據(jù)準(zhǔn)備:訓(xùn)練集、驗(yàn)證集、測試集“““
(x_train?y_train)?(x_test?y_test)?=?mnist.load_data()
x_train?=?x_train.reshape((-1?28?28?1))
x_train?=?x_train.astype(‘float32‘)/255
y_train?=?to_categorical(y_train)
x_test?=?x_test.reshape((-1?28?28?1))
x_test?=?x_test.astype(‘float32‘)/255
y_test?=?to_categorical(y_test)
##?訓(xùn)練集的前10000個(gè)樣本劃分為驗(yàn)證集
x_val?=?x_train[:10000]
y_val?=?y_train[:10000]
partial_x_train?=?x_train[10000:]
partial_y_train?=?y_train[10000:]
“““網(wǎng)絡(luò)設(shè)計(jì)“““
network?=?Sequential()
network.add(Conv2D(32?(3?3)?padding=‘same‘?activation=‘relu‘?input
評(píng)論
共有 條評(píng)論