用TensorFlow搭建神经网络

最新推荐文章于 2025-10-27 15:09:36 发布

原创

最新推荐文章于 2025-10-27 15:09:36 发布 · 671 阅读

1 ·

CC 4.0 BY-SA版权

文章标签：

#神经网络 #深度学习 #tensorflow #可视化

这篇博客详细介绍了如何使用tf.keras构建不同类型的神经网络，包括线性回归、多层神经网络（回归、二分类、多分类）以及利用函数式API实现多输入模型。同时，还讲解了在TensorFlow 1.0版本中使用tf.Session()进行会话管理、变量与placeholder操作，并通过tensorboard进行神经网络的可视化。

import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
%matplotlib

1 tf.keras

import tensorflow as tf
tf.__version__  # 2.2.0版本

1.1 tf.keras 实现线性回归

import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

data = pd.read_csv('dataset/income.csv')
sns.lmplot(x='Education', y='Income', data=data)
x = data.Education
y = data.Income

model = tf.keras.Sequential()
model.add(
	tf.keras.layers.Dense(1, input_shape=(1,))
)
model.summary()

model.compile(optimizer='adam', loss='mse')
model.fit(x, y, epochs=500)

1.2 多层神经网络——回归

from sklearn.datasets import load_boston
import pandas as pd

boston = load_boston()
data = pd.DataFrame(boston.data, columns=boston.feature_names)
data['PRICE'] = boston.target
x = data.iloc[:, :-1]
y = data.iloc[:, -1]

model = tf.keras.Sequential([
    tf.keras.layers.Dense(10, input_shape=(13,), activation='relu'), 
    tf.keras.layers.Dense(1)
])
model.summary()

model.compile(optimizer='adam', loss='mse')
model.fit(x, y, epochs=1000)

1.3 多层神经网络——二分类

from sklearn.datasets import load_breast_cancer

breast_cancer = load_breast_cancer()

data = pd.DataFrame(breast_cancer.data, columns=breast_cancer.feature_names)
data['label'] = breast_cancer.target
X = data.iloc[:, :-1]
y = data.iloc[:,-1]
y.value_counts()

model = tf.keras.Sequential([
    tf.keras.layers.Dense(10, input_shape=(30,), activation='relu'),
    tf.keras.layers.Dense(10, activation='relu'),
    tf.keras.layers.Dense(1, activation='sigmoid')
])
model.summary()

model.compile(optimizer='adam', 
              loss='binary_crossentropy',
              metrics=['accuracy'])
history = model.fit(X, y, epochs=100)

history.history.keys()  # >> dict_keys(['loss', 'accuracy'])

import matplotlib.pyplot as plt

plt.plot(history.history.get('loss'))
# plt.plot(history.history.get('accuracy'))

1.4 多层神经网络——多分类

# fashion_mnist数据集

fashion_mnist = tf.keras.datasets.fashion_mnist.load_data()

(X_train, y_train), (X_test, y_test) = fashion_mnist
X_train.shape, y_train.shape

# 查看随机10个样本图片
import matplotlib

最低0.47元/天解锁文章