前馈全连接神经网络对鸢尾花数据集进行分类

鸢尾花数据集通过from sklearn.datasets import load_iris 方式获取，并将获取的数据划分为训练集、验证集、测试集

import numpy as np

import pandas as pd

import matplotlib.pyplot as plt

from sklearn.datasets import load_iris

from sklearn.model_selection import train_test_split

iris = load_iris()

x_train,x_test,y_train,y_test=train_test_split(iris.data,iris.target,test_size=0.2,random_state=23)

X_train,X_valid,y_train,y_valid=train_test_split(x_train,y_train,test_size=0.2,random_state=12)

打印显示

print(X_valid.shape)

print(X_train.shape)

构建一个10层网络，隐藏层每层对应16个神经元，激活函数都是relu函数，输入输出神经元判断

import tensorflow as tf

from tensorflow import keras

model = keras.models.Sequential([

    keras.layers.Flatten(input_shape=[4]),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dense(16, activation="relu"),

    keras.layers.Dropout(rate=0.2),

    keras.layers.Dense(3, activation="softmax")

])

model.summary()

model.layers[1]

weights_l,bias_l=model.layers[1].get_weights()

print(weights_l.shape)

print(bias_l.shape)

编译训练网络50次

model.compile(loss="sparse_categorical_crossentropy",

             optimizer="sgd",metrics=["accuracy"])

h=model.fit(X_train,y_train,batch_size=10,epochs=50,validation_data=(X_valid,y_valid))

使用pd.DataFrame()函数，将这个对象转换为一个pandas DataFrame，这样就可以更方便地查看和分析训练过程中的指标变化。

pd.DataFrame(h.history)

绘制画图，model.evaluate()函数是用来评估一个训练好的模型在测试数据集上的性能。

pd.DataFrame(h.history).plot(figsize=(8,5))

plt.grid(True)

plt.gca().set_ylim(0,1)

plt.show()

model.evaluate(x_test,y_test,batch_size=1)