人工智能竞赛-房价预测

本文为看雪论坛优秀‍‍‍文章
看雪论坛作者ID：pureGavin

比赛网址：https://www.kaggle.com/c/house-prices-advanced-regression-techniques

本次的内容其实和泰坦尼克的内容没有本质的区别，都是二分类任务.

我的代码

# This Python 3 environment comes with many helpful analytics libraries installed
# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python
# For example, here's several helpful packages to load
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
from sklearn.ensemble import RandomForestRegressor
# Input data files are available in the read-only "../input/" directory
# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
    for filename in filenames:
        print(os.path.join(dirname, filename))
# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using "Save & Run All"
# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session
 
# 载入训练数据
train=pd.read_csv('/kaggle/input/house-prices-advanced-regression-techniques/train.csv')
train.head()
 
# 选取部分训练数据
train_y=train.SalePrice
predict_data=['LotArea','OverallQual','YearBuilt','YearRemodAdd','GrLivArea','TotRmsAbvGrd']
train_X=train[predict_data]
 
# 进行训练
model=RandomForestRegressor()
model.fit(train_X,train_y)
 
# 载入测试数据，并进行预测
test=pd.read_csv('/kaggle/input/house-prices-advanced-regression-techniques/test.csv')
test_X=test[predict_data]
predict_price=model.predict(test_X)
print(predict_price)
 
# 输出成表格的形式并提交
my_submission=pd.DataFrame({'Id':test.Id,"SalePrice":predict_price})
#my_submission.head()
my_submission.to_csv("submission.csv",index=False)

提交后查看一下分数：

分数并不是很好（0为最高分）依旧需要参考一下大神的代码。

大神的代码

#invite people for the Kaggle party
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np
from scipy.stats import norm
from sklearn.preprocessing import StandardScaler
from scipy import stats
import warnings
warnings.filterwarnings('ignore')
%matplotlib inline
#bring in the six packs
df_train = pd.read_csv('./input/train.csv')
#check the decoration
df_train.columns

前面都是一样的，加载一些库和训练文件。

# 打印出saleprice的分布
df_train['SalePrice'].describe()
# 使用图表打印出saleprice的分布情况
sns.distplot(df_train['SalePrice']);
#skewness and kurtosis
print("Skewness: %f" % df_train['SalePrice'].skew())
print("Kurtosis: %f" % df_train['SalePrice'].kurt())

# 检查grlivarea与saleprice的关系，这两组数值有线性关系
var = 'GrLivArea'
data = pd.concat([df_train['SalePrice'], df_train[var]], axis=1)
data.plot.scatter(x=var, y='SalePrice', ylim=(0,800000));

# 检查totalbsmtsf与saleprice的关系，这两组数据也有线性关系，但在某些情况下totalbsmtsf会直接归零
var = 'TotalBsmtSF'
data = pd.concat([df_train['SalePrice'], df_train[var]], axis=1)
data.plot.scatter(x=var, y='SalePrice', ylim=(0,800000));

# 检查overallqual与saleprice的关系，也是线性关系，但每个阶段都有异常值
# 分析这两组数据之间的关系用的是箱型图
var = 'OverallQual'
data = pd.concat([df_train['SalePrice'], df_train[var]], axis=1)
f, ax = plt.subplots(figsize=(8, 6))
fig = sns.boxplot(x=var, y="SalePrice", data=data)
fig.axis(ymin=0, ymax=800000);

# 建造年份与售价的关系图，可见并没有多少关系，并且这组走向数据并没有考虑到通货膨胀的问题
var = 'YearBuilt'
data = pd.concat([df_train['SalePrice'], df_train[var]], axis=1)
f, ax = plt.subplots(figsize=(16, 8))
fig = sns.boxplot(x=var, y="SalePrice", data=data)
fig.axis(ymin=0, ymax=800000);
plt.xticks(rotation=90);

# 上面我们只选择了几个关联的特征，然而这么做是不严谨的，因为这样就会把潜在可能影响saleprice的特征全部忽略
# 热力图用来表示数据与数据之间的关系，这张热力图包含了训练数据中所有的项目，并且颜色越浅表示关联越大
corrmat = df_train.corr()
f, ax = plt.subplots(figsize=(15, 10))
sns.heatmap(corrmat, vmax=.8, square=True,linewidths=0.2);

# 将上面的热力图中的数据进行筛选，选出十个与售价关系最大的特征并重新组成一张热力图
k = 10 #number of variables for heatmap
cols = corrmat.nlargest(k, 'SalePrice')['SalePrice'].index
cm = np.corrcoef(df_train[cols].values.T)
sns.set(font_scale=1.25)
hm = sns.heatmap(cm, cbar=True, annot=True, square=True, fmt='.2f',
annot_kws={'size': 10}, yticklabels=cols.values, xticklabels=cols.values,linewidths=0.2)
plt.show()

# 将上面的热力图改为散点图
sns.set()
cols = ['SalePrice', 'OverallQual', 'GrLivArea', 'GarageCars', 'TotalBsmtSF', 'FullBath', 'YearBuilt']
sns.pairplot(df_train[cols], size = 2.5)
plt.show();

# 打印出数据缺失的情况
total = df_train.isnull().sum().sort_values(ascending=False)
#print(total.head(20))
percent = (df_train.isnull().sum()/df_train.isnull().count()).sort_values(ascending=False)
missing_data = pd.concat([total, percent], axis=1, keys=['Total', 'Percent'])
missing_data.head(20)

# 处理缺失的数据
df_train = df_train.drop((missing_data[missing_data['Total'] > 1]).index,1)
df_train = df_train.drop(df_train.loc[df_train['Electrical'].isnull()].index)
# 检查是否还有缺失的数据
df_train.isnull().sum().max()

# 对数据中的离群值进行处理，低范围的值都差不多，但是高范围的值相差很多，需要着重注意一下最高的两个7
saleprice_scaled = StandardScaler().fit_transform(df_train['SalePrice'][:,np.newaxis]);
low_range = saleprice_scaled[saleprice_scaled[:,0].argsort()][:10]
high_range= saleprice_scaled[saleprice_scaled[:,0].argsort()][-10:]
print('outer range (low) of the distribution:')
print(low_range)
print('\nouter range (high) of the distribution:')
print(high_range)

# 观察grlivarea与saleprice的关系，可以看到grlivarea有两个特别大的值但是saleprice都不高，应当认定为离群值，需要删除
# 但是顶部的两个值虽然看似离群值，但是是顺应趋势的，需要保留
var = 'GrLivArea'
data = pd.concat([df_train['SalePrice'], df_train[var]], axis=1)
data.plot.scatter(x=var, y='SalePrice', ylim=(0,800000));

# 删除grlivarea中的离群值
df_train.sort_values(by = 'GrLivArea', ascending = False)[:2]
df_train = df_train.drop(df_train[df_train['Id'] == 1299].index)
df_train = df_train.drop(df_train[df_train['Id'] == 524].index)
 
# 打印totalbsmtsf与saleprice的关系，有三个大于三千的值，但是不需要处理，对于离群值的处理需要谨慎，并不是看到就删除
var = 'TotalBsmtSF'
data = pd.concat([df_train['SalePrice'], df_train[var]], axis=1)
data.plot.scatter(x=var, y='SalePrice', ylim=(0,800000));

# 打印出直方图和正态概率图
sns.distplot(df_train['SalePrice'], fit=norm);
fig = plt.figure()
res = stats.probplot(df_train['SalePrice'], plot=plt)

# 应用对数变换
df_train['SalePrice'] = np.log(df_train['SalePrice'])
 
# 变换后的直方图和正态概率图
sns.distplot(df_train['SalePrice'], fit=norm);
fig = plt.figure()
res = stats.probplot(df_train['SalePrice'], plot=plt)

# grlivarea在对数变换前的直方图和正态概率图
sns.distplot(df_train['GrLivArea'], fit=norm);
fig = plt.figure()
res = stats.probplot(df_train['GrLivArea'], plot=plt)

# 应用对数变换
df_train['GrLivArea'] = np.log(df_train['GrLivArea'])
 
# grlivarea做了对数变换后的直方图和正态概率图
sns.distplot(df_train['GrLivArea'], fit=norm);
fig = plt.figure()
res = stats.probplot(df_train['GrLivArea'], plot=plt)

# totalbsmtsf在做对数转换前的直方图和正态概率图
sns.distplot(df_train['TotalBsmtSF'], fit=norm);
fig = plt.figure()
res = stats.probplot(df_train['TotalBsmtSF'], plot=plt)

# totalbsmtsf这项数据有些特殊，因为有很多0值，而0是没有对数的，所以不能进行对数转换
# 但是有许多非零的值是可以转换的，所以我们需要暂时忽略零值
df_train['HasBsmt'] = pd.Series(len(df_train['TotalBsmtSF']), index=df_train.index)
df_train['HasBsmt'] = 0
df_train.loc[df_train['TotalBsmtSF']>0,'HasBsmt'] = 1
 
# 应用对数变换
df_train.loc[df_train['HasBsmt']==1,'TotalBsmtSF'] = np.log(df_train['TotalBsmtSF'])
 
# 转换后的方直图和正态概率图
sns.distplot(df_train[df_train['TotalBsmtSF']>0]['TotalBsmtSF'], fit=norm);
fig = plt.figure()
res = stats.probplot(df_train[df_train['TotalBsmtSF']>0]['TotalBsmtSF'], plot=plt)

# 之前的grlivarea与saleprice的图是类似于圆锥的形状（同质问题严重）
# 做了正态处理后，散点图就不再有圆锥型状了（解决了同质问题）
plt.scatter(df_train['GrLivArea'], df_train['SalePrice']);

# 对totalbsmtsf做正态处理后，同质问题也得到了解决
plt.scatter(df_train[df_train['TotalBsmtSF']>0]['TotalBsmtSF'], df_train[df_train['TotalBsmtSF']>0]['SalePrice']);

# 将分类变量转换为虚拟变量
df_train = pd.get_dummies(df_train)

剩下的就和我自己的代码一样了，就是用随机森林做二分类判断，然后将训练好的模型直接对test.csv进行预测，然后将预测好的结果输出到submission.csv，这里不知道为什么没有写下去。

诸如此类二分类任务都是先分析每个特征和要预测的值之间的关系，然后对特征进行增强，需要注意的是泰坦尼克和本次的房价预测用的都是随机森林（线性），还有用tanh或者relu（逻辑）的情况同样值得学习。

看雪ID：pureGavin

https://bbs.pediy.com/user-home-777502.htm

*本文由看雪论坛 pureGavin 原创，转载请注明来自看雪社区

# 往期推荐

球分享

球点赞

球在看

点击“阅读原文”，了解更多！