15.手写数字识别-小数据集

1.手写数字数据集

• from sklearn.datasets import load_digits
• digits = load_digits()

代码如下：

from sklearn.datasets import load_digits
import numpy as np

# 1.手写数字数据集
digits = load_digits()
X_data = digits.data.astype(np.float32)
Y_data = digits.target.astype(np.float32).reshape(-1, 1)  # 将Y_data变为一列
print("X_data:
", X_data, "
Y_data:
", Y_data)

运行结果图如下：

2.图片数据预处理

• x：归一化MinMaxScaler()
• y：独热编码OneHotEncoder()或to_categorical
• 训练集测试集划分
• 张量结构

 代码如下：

# 2.图片数据预处理
# 1）归一化
from sklearn.preprocessing import MinMaxScaler
from sklearn.preprocessing import OneHotEncoder
from sklearn.model_selection import train_test_split

scaler = MinMaxScaler()
X_data = scaler.fit_transform(X_data)
print('MinMaxScaler_trans_X_data:')
print(X_data)

# 2）独热编码
Y = OneHotEncoder().fit_transform(Y_data).todense()   # one-hot编码
print('OneHot_Y:')
print(Y)

# 3）划分训练集测试集
X = X_data.reshape(-1, 8, 8, 1)        # 张量结构
X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, random_state=0, stratify=Y)
print("X.shape:
", X.shape,  "
X_train.shape:
", X_train.shape, "
X_test.shape:
",
      X_test.shape, "
Y_train.shape:
", Y_train.shape, "
Y_test.shape:
", Y_test.shape)

运行结果图如下：

3.设计卷积神经网络结构

• 绘制模型结构图，并说明设计依据。

 代码如下：

# 3.设计卷积神经网络结构
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPool2D

# 建立模型
model = Sequential()
# 一层卷积
model.add(Conv2D(filters=16,
                 kernel_size=(5, 5),
                 padding='same',
                 input_shape=X_train.shape[1:],
                 activation='relu'))
# 池化层1
model.add(MaxPool2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
# 二层卷积
model.add(Conv2D(filters=32,
                 kernel_size=(5, 5),
                 padding='same',
                 activation='relu'))
# 池化层2
model.add(MaxPool2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
# 三层卷积
model.add(Conv2D(filters=64,
                 kernel_size=(5, 5),
                 padding='same',
                 activation='relu'))
# 四层卷积
model.add(Conv2D(filters=128,
                 kernel_size=(5, 5),
                 padding='same',
                 activation='relu'))
# 池化层3
model.add(MaxPool2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
# 平坦层
model.add(Flatten())
# 全连接层
model.add(Dense(128, activation='relu'))
model.add(Dropout(0.25))
# 激活函数
model.add(Dense(10, activation='softmax'))
print("变化过程:
")
model.summary()

运行结果图如下：

4.模型训练

• model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
• train_history = model.fit(x=X_train,y=y_train,validation_split=0.2, batch_size=300,epochs=10,verbose=2)

代码如下：

# 4.模型训练
model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
train_history = model.fit(x=X_train, y=Y_train, validation_split=0.2, batch_size=300, epochs=55, verbose=2)

运行结果图如下：

 

5.模型评价

• model.evaluate()
• 交叉表与交叉矩阵
• pandas.crosstab
• seaborn.heatmap

代码如下：

# 5.模型评价
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt


score = model.evaluate(X_test, Y_test)
print(score)

# 交叉表与交叉矩阵
# 1)预测值
y_pred = model.predict_classes(X_test)
print(y_pred[:10])

# 2)交叉表查看预测数据与元数据对比
y_test1 = np.argmax(Y_test, axis=1).reshape(-1)
y_true = np.array(y_test1)[0]
print(y_test1)
pd.crosstab(y_true, y_pred, rownames=["true"], colnames=["predict"])

# 3)交叉矩阵
a = pd.crosstab(np.array(y_test1).reshape(-1), y_pred)
df = pd.DataFrame(a)
sns.heatmap(df, annot=True, cmap='summer', linewidths=0.2, linecolor='G')
plt.show()

运行结果图如下：