Pytorch：全连接神经网络-MLP回归

Pytorch: 全连接神经网络-解决 Boston 房价回归问题

Copyright: Jingmin Wei, Pattern Recognition and Intelligent System, School of Artificial and Intelligence, Huazhong University of Science and Technology

Pytorch教程专栏链接

MLP 回归模型

使用sklearn库的fetch_california_housing()函数。数据集共包含20640个样本，有8个自变量。

import numpy as np
import pandas as pd
from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error, mean_absolute_error
from sklearn.datasets import fetch_california_housing

import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import SGD
import torch.utils.data as Data
import matplotlib.pyplot as plt
import seaborn as sns

房价数据准备

# 导入数据
housedata = fetch_california_housing()
# 切分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(housedata.data, housedata.target,
                                                    test_size = 0.3, random_state = 42)

70% 训练集，30%测试集。

X_train, X_test, y_train, y_test

(array([[   4.1312    ,   35.        ,    5.88235294, ...,    2.98529412,
           33.93      , -118.02      ],
        [   2.8631    ,   20.        ,    4.40120968, ...,    2.0141129 ,
           32.79      , -117.09      ],
        [   4.2026    ,   24.        ,    5.61754386, ...,    2.56491228,
           34.59      , -120.14      ],
        ...,
        [   2.9344    ,   36.        ,    3.98671727, ...,    3.33206831,
           34.03      , -118.38      ],
        [   5.7192    ,   15.        ,    6.39534884, ...,    3.17889088,
           37.58      , -121.96      ],
        [   2.5755    ,   52.        ,    3.40257649, ...,    2.10869565,
           37.77      , -122.42      ]]),
 array([[   1.6812    ,   25.        ,    4.19220056, ...,    3.87743733,
           36.06      , -119.01      ],
        [   2.5313    ,   30.        ,    5.03938356, ...,    2.67979452,
           35.14      , -119.46      ],
        [   3.4801    ,   52.        ,    3.97715472, ...,    1.36033229,
           37.8       , -122.44      ],
        ...,
        [   3.512     ,   16.        ,    3.76228733, ...,    2.36956522,
           33.67      , -117.91      ],
        [   3.65      ,   10.        ,    5.50209205, ...,    3.54751943,
           37.82      , -121.28      ],
        [   3.052     ,   17.        ,    3.35578145, ...,    2.61499365,
           34.15      , -118.24      ]]),
 array([1.938, 1.697, 2.598, ..., 2.221, 2.835, 3.25 ]),
 array([0.477  , 0.458  , 5.00001, ..., 2.184  , 1.194  , 2.098  ]))

# 数据标准化处理
scale = StandardScaler()
X_train_s = scale.fit_transform(X_train)
X_test_s = scale.transform(X_test)

# 将训练数据转为数据表
housedatadf = pd.DataFrame(data=X_train_s, columns = housedata.feature_names)
housedatadf['target'] = y_train
housedatadf