Keras (3) implements a deep neural network

This article will add a DNN deep learning model based on the previous section.

One, print the version information of the python library used

import matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np
import sklearn
import pandas as pd
import os
import sys
import time
import tensorflow as tf
from tensorflow import keras

# 1,打印使用的python库的版本信息
print(sys.version_info)
for module in mpl, np, pd, sklearn, tf, keras:
    print(module.__name__, module.__version__)

#---output-----------
sys.version_info(major=3, minor=8, micro=5, releaselevel='final', serial=0)
matplotlib 3.3.2
numpy 1.19.1
pandas 1.0.3
sklearn 0.23.2
tensorflow 2.2.0
tensorflow.keras 2.3.0-tf

Second, extract the image data set-"training data", "validation data", "test data" from keras.datasets, and print its dimensions

fashion_mnist = keras.datasets.fashion_mnist
(x_train_all, y_train_all), (x_test, y_test) = fashion_mnist.load_data()
x_valid, x_train = x_train_all[:5000], x_train_all[5000:]
y_valid, y_train = y_train_all[:5000], y_train_all[5000:]

print(x_valid.shape, y_valid.shape)
print(x_train.shape, y_train.shape)
print(x_test.shape, y_test.shape)

#---------output--------------------
(5000, 28, 28) (5000,)
(55000, 28, 28) (55000,)
(10000, 28, 28) (10000,)

3. Standardize data processing

# x = (x - u) / std
from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()
# x_train: [None, 28, 28] -> [None, 784]
x_train_scaled = scaler.fit_transform(
    x_train.astype(np.float32).reshape(-1, 1)).reshape(-1, 28, 28)
x_valid_scaled = scaler.transform(
    x_valid.astype(np.float32).reshape(-1, 1)).reshape(-1, 28, 28)
x_test_scaled = scaler.transform(
    x_test.astype(np.float32).reshape(-1, 1)).reshape(-1, 28, 28)

Fourth, build DNN deep neural network model

# tf.keras.models.Sequential()
model = keras.models.Sequential()
model.add(keras.layers.Flatten(input_shape=[28, 28]))
for _ in range(50):
    model.add(keras.layers.Dense(100, activation="relu"))
model.add(keras.layers.Dense(10, activation="softmax"))

model.compile(loss="sparse_categorical_crossentropy",
              optimizer = keras.optimizers.SGD(0.01),
              metrics = ["accuracy"])

Five, view the overall structure of the model-and parameters

print(model.summary())

Six, add callback function, training model

# Tensorboard, earlystopping, ModelCheckpoint
logdir = './dnn-callbacks'
if not os.path.exists(logdir):
    os.mkdir(logdir)
output_model_file = os.path.join(logdir,"fashion_mnist_model.h5")

callbacks = [
    keras.callbacks.TensorBoard(logdir),
    keras.callbacks.ModelCheckpoint(output_model_file,save_best_only = True),
    keras.callbacks.EarlyStopping(patience=10, min_delta=1e-3),
]
history = model.fit(x_train_scaled, y_train, epochs=20,
                    validation_data=(x_valid_scaled, y_valid),
                    callbacks = callbacks)

Seven, drawing the process of change in the training model

During the training process, the gradient may explode or disappear. For specific reasons, please refer to: Detailed explanation of gradient disappearance, explosion reasons and solutions in machine learning.
Here is the link to the author’s detailed derivation and extreme value process of the sigmoid function as follows:
sigmoid function derivation, extreme value (the most detailed in history )

def plot_learning_curves(history):
    pd.DataFrame(history.history).plot(figsize=(8, 5))
    plt.grid(True)
    plt.gca().set_ylim(0, 3)
    plt.show()

plot_learning_curves(history)

Eight, call the estimator

print(model.evaluate(x_test_scaled, y_test, verbose=0))

Nine, code summary

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

import matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np
import sklearn
import pandas as pd
import os
import sys
import time
import tensorflow as tf
from tensorflow import keras

# 1,打印使用的python库的版本信息
print(sys.version_info)
for module in mpl, np, pd, sklearn, tf, keras:
    print(module.__name__, module.__version__)


# 2,从keras.datasets中提取图片数据集-“训练数据”、“验证数据”、“测试数据”，并打印其维度
fashion_mnist = keras.datasets.fashion_mnist
(x_train_all, y_train_all), (x_test, y_test) = fashion_mnist.load_data()
x_valid, x_train = x_train_all[:5000], x_train_all[5000:]
y_valid, y_train = y_train_all[:5000], y_train_all[5000:]

print(x_valid.shape, y_valid.shape)
print(x_train.shape, y_train.shape)
print(x_test.shape, y_test.shape)

# 3,对数据标准化处理
# x = (x - u) / std
from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()
# x_train: [None, 28, 28] -> [None, 784]
x_train_scaled = scaler.fit_transform(
    x_train.astype(np.float32).reshape(-1, 1)).reshape(-1, 28, 28)
x_valid_scaled = scaler.transform(
    x_valid.astype(np.float32).reshape(-1, 1)).reshape(-1, 28, 28)
x_test_scaled = scaler.transform(
    x_test.astype(np.float32).reshape(-1, 1)).reshape(-1, 28, 28)

# 4,构建DNN深度神经网络模型
# tf.keras.models.Sequential()
model = keras.models.Sequential()
model.add(keras.layers.Flatten(input_shape=[28, 28]))
for _ in range(30):
    model.add(keras.layers.Dense(100, activation="relu"))
model.add(keras.layers.Dense(10, activation="softmax"))

model.compile(loss="sparse_categorical_crossentropy",
              optimizer = keras.optimizers.SGD(0.01),
              metrics = ["accuracy"])

# 5,查看模型整体结构-及参数
print(model.summary())

# 6,添加回调函数、训练模型
# Tensorboard, earlystopping, ModelCheckpoint
logdir = './dnn-callbacks'
if not os.path.exists(logdir):
    os.mkdir(logdir)
output_model_file = os.path.join(logdir,"fashion_mnist_model.h5")

callbacks = [
    keras.callbacks.TensorBoard(logdir),
    keras.callbacks.ModelCheckpoint(output_model_file,save_best_only = True),
    keras.callbacks.EarlyStopping(patience=10, min_delta=1e-3),
]
history = model.fit(x_train_scaled, y_train, epochs=20,
                    validation_data=(x_valid_scaled, y_valid),
                    callbacks = callbacks)

# 7,绘制训练模型中，变化过程
def plot_learning_curves(history):
    pd.DataFrame(history.history).plot(figsize=(8, 5))
    plt.grid(True)
    plt.gca().set_ylim(0, 3)
    plt.show()

plot_learning_curves(history)

# 1. 参数众多，训练不充分
# 2. 梯度消失 -> 链式法则 -> 复合函数f(g(x))

# 8,调用估计器
print(model.evaluate(x_test_scaled, y_test, verbose=0))