【深度学习】用Sequential顺序模型(tensorflow)进行房价预测

数据选用 

https://archive.ics.uci.edu/ml/machine-learning-databases/housing/housing.datahttps://archive.ics.uci.edu/ml/machine-learning-databases/housing/housing.data

代码

数据处理

f = open(r"housing.data.txt")
lines = f.readlines()
with open("data.csv","w+") as t:

    for i in lines:
        i = ",".join(i.split())
        t.writelines([i])
        t.write("\n")

模型构建

import numpy as np
import pandas as pd
import tensorflow as tf
from keras.layers import Dense
import matplotlib.pyplot as plt


# 划分数据集
train_data = data[:int(len(data)*0.9)]
y_train = train_data[:,-1]

test_data = data[int(len(data)*0.9):]
y_test = test_data[:,-1]


# 添加序列模型
model = tf.keras.Sequential([
    Dense(128,input_shape=(14,),activation="relu"),
    Dense(64,activation="relu"),
    Dense(1)
])
model.compile(optimizer="adam",loss="mse")
model.summary()


# 对模型进行拟合
model.fit(train_data,y_train,epochs=10,batch_size=8)

# 对模型进行评估 0.23200644552707672
model.evaluate(test_data,y_test)

# 利用测试集对数据进行预测
result = model.predict(test_data)

结果可视化

new_data = pd.DataFrame({"Prediction":result.reshape(-1),"Actual":y_test.reshape(-1)})
new_data.plot()
plt.show()

回归算法

数据选用：

链接：百度网盘 请输入提取码百度网盘为您提供文件的网络备份、同步和分享服务。空间大、速度快、安全稳固，支持教育网加速，支持手机端。注册使用百度网盘即可享受免费存储空间https://pan.baidu.com/s/1WELkePK7KCKFtrrghguueg?pwd=n73p

利用机器学习（LinearRegression）方法对比预测结果

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error 
from sklearn.model_selection import train_test_split


data = pd.read_csv("data.csv")

# 数据划分
X,y = np.array(data.drop(["MEDV"],1)),np.array(data['MEDV'])

# 训练集和测试集划分
X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.1) # random_state=0

# 模型拟合
linear = LinearRegression(n_jobs=-1)
linear.fit(X_train, y_train)

# 评估 0.8124025816789091
acc = linear.score(X_test, y_test)

prediction = linear.predict(X_test)

# 计算预测和实际值均方差 21.916277197562923
mse = mean_squared_error(y_test, prediction)

# 可视化
contrast = pd.DataFrame({"Prediction":prediction,"Acutal":y_test})
contrast.plot()
plt.show()