Kaggle: House Prices: Advanced Regression Techniques

notebook來自https://www.kaggle.com/neviadomski/how-to-get-to-top-25-with-simple-model-sklearn

思路流程：

1.導(dǎo)入數(shù)據(jù)，查看數(shù)據(jù)結(jié)構(gòu)和缺失值情況
重點在于查看缺失值情況的寫法:
NAs = pd.concat([train.isnull().sum(), test.isnull().sum()], axis = 1, keys = ['train', 'test']) NAs[NAs.sum(axis=1) > 0]

2.數(shù)據(jù)預(yù)處理（刪除無用特征，特征轉(zhuǎn)化，缺失值填充，構(gòu)造新特征，特征值標(biāo)準(zhǔn)化，轉(zhuǎn)化為dummy）
Q:什么樣的特征需要做轉(zhuǎn)化？
A:如某些整型數(shù)據(jù)只表示類別，其數(shù)值本身沒有意義，則應(yīng)轉(zhuǎn)化為dummy
重點學(xué)習(xí)手動將特征轉(zhuǎn)化為dummy的方法（這里情況稍微還要復(fù)雜一點，因為存在同一特征對應(yīng)兩列的情況，如Condition1,Condition2)

3.隨機打亂數(shù)據(jù)，分離訓(xùn)練集和測試集

4.構(gòu)建多個單一模型

5.模型融合

問題：

1.如何判斷一個特征是否是無用特征？

2.模型融合的方法？這里為什是np.exp(GB_model.predict(test_features)) + np.exp(ENS_model.predict(test_features_std))？

3.為什么label分布偏斜需要做轉(zhuǎn)化？

In?[33]: #Kaggle: House Prices: Advanced Regression Techniques import numpy as np import pandas as pd import matplotlib.pyplot as plt import seaborn as sns from sklearn import ensemble, linear_model, tree from sklearn.model_selection import train_test_split, cross_val_score from sklearn.metrics import mean_squared_error, r2_score from sklearn.utils import shuffle ? %matplotlib inline import warnings warnings.filterwarnings('ignore') ? train = pd.read_csv('downloads/train.csv') test = pd.read_csv('downloads/test.csv') In?[8]: train.head() Out[8]: ?IdMSSubClassMSZoningLotFrontageLotAreaStreetAlleyLotShapeLandContourUtilities...PoolAreaPoolQCFenceMiscFeatureMiscValMoSoldYrSoldSaleTypeSaleConditionSalePrice01234

1	60	RL	65.0	8450	Pave	NaN	Reg	Lvl	AllPub	...	0	NaN	NaN	NaN	0	2	2008	WD	Normal	208500
2	20	RL	80.0	9600	Pave	NaN	Reg	Lvl	AllPub	...	0	NaN	NaN	NaN	0	5	2007	WD	Normal	181500
3	60	RL	68.0	11250	Pave	NaN	IR1	Lvl	AllPub	...	0	NaN	NaN	NaN	0	9	2008	WD	Normal	223500
4	70	RL	60.0	9550	Pave	NaN	IR1	Lvl	AllPub	...	0	NaN	NaN	NaN	0	2	2006	WD	Abnorml	140000
5	60	RL	84.0	14260	Pave	NaN	IR1	Lvl	AllPub	...	0	NaN	NaN	NaN	0	12	2008	WD	Normal	250000

5 rows × 81 columns?

In?[9]: #檢查缺失值 NAs = pd.concat([train.isnull().sum(), test.isnull().sum()], axis = 1, keys = ['train', 'test']) #sum()默認(rèn)的axis=0，即跨行 NAs[NAs.sum(axis=1) > 0] #只顯示有缺失值的特征 Out[9]: ?traintestAlleyBsmtCondBsmtExposureBsmtFinSF1BsmtFinSF2BsmtFinType1BsmtFinType2BsmtFullBathBsmtHalfBathBsmtQualBsmtUnfSFElectricalExterior1stExterior2ndFenceFireplaceQuFunctionalGarageAreaGarageCarsGarageCondGarageFinishGarageQualGarageTypeGarageYrBltKitchenQualLotFrontageMSZoningMasVnrAreaMasVnrTypeMiscFeaturePoolQCSaleTypeTotalBsmtSFUtilities

1369	1352.0
37	45.0
38	44.0
0	1.0
0	1.0
37	42.0
38	42.0
0	2.0
0	2.0
37	44.0
0	1.0
1	0.0
0	1.0
0	1.0
1179	1169.0
690	730.0
0	2.0
0	1.0
0	1.0
81	78.0
81	78.0
81	78.0
81	76.0
81	78.0
0	1.0
259	227.0
0	4.0
8	15.0
8	16.0
1406	1408.0
1453	1456.0
0	1.0
0	1.0
0	2.0

In?[10]: #打印R2和RMSE得分 def print_score (prediction, labels):print('R2: {}'.format(r2_score(prediction, labels)))print('RMSE: {}'.format(np.sqrt(mean_squared_error(prediction, labels)))) ? #對給定的模型進行評估，分別打印訓(xùn)練集上的得分和測試集上的得分 def train_test_score(estimator, x_train, x_test, y_train, y_test):train_predictions = estimator.predict(x_train)print('------------train-----------')print_score(train_predictions, y_train)print('------------test------------')test_predictions = estimator.predict(x_test)print_score(test_predictions, y_test) In?[11]: #將標(biāo)簽從訓(xùn)練集中分離出來 train_label = train.pop('SalePrice') ? #將訓(xùn)練集特征和測試集特征拼在一起，便于一起刪除無用的特征 features = pd.concat([train, test], keys = ['train', 'test']) ? #刪除無用特征（為什么說它們是無用特征并沒有解釋） features.drop(['Utilities', 'RoofMatl', 'MasVnrArea', 'BsmtFinSF1', 'BsmtFinSF2', 'BsmtUnfSF', 'Heating', 'LowQualFinSF','BsmtFullBath', 'BsmtHalfBath', 'Functional', 'GarageYrBlt', 'GarageArea', 'GarageCond', 'WoodDeckSF','OpenPorchSF', 'EnclosedPorch', '3SsnPorch', 'ScreenPorch', 'PoolArea', 'PoolQC', 'Fence', 'MiscFeature', 'MiscVal'],axis=1, inplace=True) print(features.shape) (2919, 56) In?[12]: #將series數(shù)據(jù)轉(zhuǎn)化為str #問題：什么樣的數(shù)據(jù)需要轉(zhuǎn)化為str #答：將原來的某些整型數(shù)據(jù)轉(zhuǎn)化為str，這些整型數(shù)據(jù)數(shù)字大小本身并沒有含義，而只是代表一個類，所以轉(zhuǎn)化為str后，后續(xù)再轉(zhuǎn)化為dummy features['MSSubClass'] = features['MSSubClass'].astype(str) #pandas調(diào)用特征的兩種方法：.feature和['feature']，兩者效果相同，下面就是.feature方法 features.OverallCond = features.OverallCond.astype(str) features['KitchenAbvGr'] = features['KitchenAbvGr'].astype(str) features['YrSold'] = features['YrSold'].astype(str) features['MoSold'] = features['MoSold'].astype(str) ? #用眾數(shù)填充缺失值 features['MSZoning'] = features['MSZoning'].fillna(features['MSZoning'].mode()[0]) features['MasVnrType'] = features['MasVnrType'].fillna(features['MasVnrType'].mode()[0]) features['Electrical'] = features['Electrical'].fillna(features['Electrical'].mode()[0]) features['KitchenQual'] = features['KitchenQual'].fillna(features['KitchenQual'].mode()[0]) features['SaleType'] = features['SaleType'].fillna(features['SaleType'].mode()[0]) ? #用某個特定值填充缺失值 features['LotFrontage'] = features['LotFrontage'].fillna(features['LotFrontage'].mean()) features['Alley'] = features['Alley'].fillna('NOACCESS') for col in ('BsmtQual', 'BsmtCond', 'BsmtExposure', 'BsmtFinType1', 'BsmtFinType2'):features[col] = features[col].fillna('NoBSMT') features['TotalBsmtSF'] = features['TotalBsmtSF'].fillna(0) features['FireplaceQu'] = features['FireplaceQu'].fillna('NoFP') for col in ('GarageType', 'GarageFinish', 'GarageQual'):features[col] = features[col].fillna('NoGRG') features['GarageCars'] = features['GarageCars'].fillna(0.0) ? #構(gòu)造新特征 features['TotalSF'] = features['TotalBsmtSF'] + features['1stFlrSF'] + features['2ndFlrSF'] features.drop(['TotalBsmtSF', '1stFlrSF', '2ndFlrSF'], axis=1, inplace=True) print(features.shape) (2919, 54) In?[13]: #查看房價分布情況 ax = sns.distplot(train_label) In?[14]: #發(fā)現(xiàn)圖像整體向左傾斜，所以做log轉(zhuǎn)變 train_label = np.log(train_label) ax = sns.distplot(train_label) In?[15]: #對數(shù)字特征做標(biāo)準(zhǔn)化處理 num_features = features.loc[:,['LotFrontage', 'LotArea', 'GrLivArea', 'TotalSF']] num_features_standarized = (num_features - num_features.mean()) / num_features.std() num_features_standarized.head() Out[15]: ??LotFrontageLotAreaGrLivAreaTotalSFtrain01234

-0.202033	-0.217841	0.413476	0.022999
0.501785	-0.072032	-0.471810	-0.029167
-0.061269	0.137173	0.563659	0.196886
-0.436639	-0.078371	0.427309	-0.092511
0.689469	0.518814	1.377806	0.988072

In?[16]: ax = sns.pairplot(num_features_standarized) In?[17]: #重點 #convert categorical data to dummies #將所有condition不重復(fù)的記錄在一個set中 conditions = set([x for x in features['Condition1']] + [x for x in features['Condition2']]) #自定義dummy變量，行數(shù)為陽歷數(shù)，列數(shù)為原condition數(shù)據(jù)轉(zhuǎn)化為dummy后的維數(shù) dummies = pd.DataFrame(data = np.zeros((len(features.index), len(conditions))), index = features.index, columns = conditions) #遍歷所有樣例，將原來的condition信息轉(zhuǎn)化為對應(yīng)的dummy信息 for i, cond in enumerate(zip(features['Condition1'], features['Condition2'])): #用ix找到位置，注意cond可能包含Condition1和Condition2兩個位置的信息，對應(yīng)dummies數(shù)組的兩個點，所以需要用ix而不能簡單的直接用dummies[i,cond]dummies.ix[i, cond] = 1 #將dummy后的特征數(shù)據(jù)拼接到原features后面，并給dummy特征的index增加前綴 features = pd.concat([features, dummies.add_prefix('Cond_')], axis = 1) #最后就可以刪除原來的Condition特征 features.drop(['Condition1', 'Condition2'], axis = 1, inplace =True) print(features.shape) (2919, 61) In?[18]: features.head() Out[18]: ??IdMSSubClassMSZoningLotFrontageLotAreaStreetAlleyLotShapeLandContourLotConfig...TotalSFCond_PosACond_ArteryCond_PosNCond_RRAnCond_RRAeCond_FeedrCond_NormCond_RRNnCond_RRNetrain01234

1	60	RL	65.0	8450	Pave	NOACCESS	Reg	Lvl	Inside	...	2566.0	0.0	0.0	0.0	0.0	0.0	0.0	1.0	0.0	0.0
2	20	RL	80.0	9600	Pave	NOACCESS	Reg	Lvl	FR2	...	2524.0	0.0	0.0	0.0	0.0	0.0	1.0	1.0	0.0	0.0
3	60	RL	68.0	11250	Pave	NOACCESS	IR1	Lvl	Inside	...	2706.0	0.0	0.0	0.0	0.0	0.0	0.0	1.0	0.0	0.0
4	70	RL	60.0	9550	Pave	NOACCESS	IR1	Lvl	Corner	...	2473.0	0.0	0.0	0.0	0.0	0.0	0.0	1.0	0.0	0.0
5	60	RL	84.0	14260	Pave	NOACCESS	IR1	Lvl	FR2	...	3343.0	0.0	0.0	0.0	0.0	0.0	0.0	1.0	0.0	0.0

5 rows × 61 columns

In?[19]: #convert Exterior to dummies Exterior = set([x for x in features['Exterior1st']] + [x for x in features['Exterior2nd']]) dummies = pd.DataFrame(data = np.zeros([len(features.index), len(Exterior)]), index = features.index, columns = Exterior) for i, ext in enumerate(zip(features['Exterior1st'], features['Exterior2nd'])):dummies.ix[i, ext] = 1 features = pd.concat([features, dummies.add_prefix('Ext_')], axis = 1) features.drop(['Exterior1st', 'Exterior2nd', 'Ext_nan'], axis = 1, inplace = True) print(features.shape) (2919, 78) In?[20]: features.dtypes[features.dtypes == 'object'].index Out[20]: Index(['MSSubClass', 'MSZoning', 'Street', 'Alley', 'LotShape', 'LandContour','LotConfig', 'LandSlope', 'Neighborhood', 'BldgType', 'HouseStyle','OverallCond', 'RoofStyle', 'MasVnrType', 'ExterQual', 'ExterCond','Foundation', 'BsmtQual', 'BsmtCond', 'BsmtExposure', 'BsmtFinType1','BsmtFinType2', 'HeatingQC', 'CentralAir', 'Electrical', 'KitchenAbvGr','KitchenQual', 'FireplaceQu', 'GarageType', 'GarageFinish','GarageQual', 'PavedDrive', 'MoSold', 'YrSold', 'SaleType','SaleCondition'],dtype='object') In?[21]: #遍歷特定類型數(shù)據(jù)的方法：for col in features.dtypes[features.dtypes == 'object'].index #convert all other categorical vars to dummies for col in features.dtypes[features.dtypes == 'object'].index:for_dummy = features.pop(col)features = pd.concat([features, pd.get_dummies(for_dummy, prefix = col)], axis = 1) print(features.shape) (2919, 263) In?[22]: #用之前幾個標(biāo)準(zhǔn)化的數(shù)據(jù)更新features ? features_standardized = features.copy() features_standardized.update(num_features_standarized) In?[23]: #重新分離訓(xùn)練集和測試集 ? #首先分離沒有標(biāo)準(zhǔn)化的features train_features = features.loc['train'].drop(['Id'], axis=1).select_dtypes(include=[np.number]).values test_features = features.loc['test'].drop(['Id'], axis=1).select_dtypes(include=[np.number]).values ? #再分離標(biāo)準(zhǔn)化的數(shù)據(jù) train_features_std = features_standardized.loc['train'].drop(['Id'], axis=1).select_dtypes(include=[np.number]).values test_features_std = features_standardized.loc['test'].drop(['Id'], axis=1).select_dtypes(include=[np.number]).values print(train_features.shape) print(train_features_std.shape) (1460, 262) (1460, 262) In?[24]: #shuffle train dataset train_features_std, train_features, train_label = shuffle(train_features_std, train_features, train_label, random_state = 5) In?[25]: #split train and test data x_train, x_test, y_train, y_test = train_test_split(train_features, train_label, test_size = 0.1, random_state = 200) x_train_std, x_test_std, y_train_std, y_test_std = train_test_split(train_features_std, train_label, test_size = 0.1, random_state = 200) In?[26]: #構(gòu)建第一個模型：ElasticNet ENSTest = linear_model.ElasticNetCV(alphas=[0.0001, 0.0005, 0.001, 0.01, 0.1, 1, 10], l1_ratio=[.01, .1, .5, .9, .99], max_iter=5000).fit(x_train_std, y_train_std) train_test_score(ENSTest, x_train_std, x_test_std, y_train_std, y_test_std) ------------train----------- R2: 0.9009283127352861 RMSE: 0.11921419084690392 ------------test------------ R2: 0.8967299522701895 RMSE: 0.11097042840114624 In?[27]: #測試模型的交叉驗證得分 score = cross_val_score(ENSTest, train_features_std, train_label, cv = 5) print('Accurary: %0.2f +/- %0.2f' % (score.mean(), score.std()*2)) Accurary: 0.88 +/- 0.10 In?[28]: #構(gòu)建第二個模型：GradientBoosting GB = ensemble.GradientBoostingRegressor(n_estimators=3000, learning_rate = 0.05, max_depth = 3, max_features = 'sqrt', min_samples_leaf = 15,min_samples_split = 10, loss = 'huber').fit(x_train_std, y_train_std) train_test_score(GB, x_train_std, x_test_std, y_train_std, y_test_std) ------------train----------- R2: 0.9607778449577035 RMSE: 0.07698826081848897 ------------test------------ R2: 0.9002871760789876 RMSE: 0.10793269100940146 In?[29]: #構(gòu)建第二個模型：GradientBoosting GB = ensemble.GradientBoostingRegressor(n_estimators=3000, learning_rate = 0.05, max_depth = 3, max_features = 'sqrt', min_samples_leaf = 15,min_samples_split = 10, loss = 'huber').fit(x_train_std, y_train_std) train_test_score(GB, x_train_std, x_test_std, y_train_std, y_test_std) Accurary: 0.90 +/- 0.04 In?[30]: #模型融合 GB_model = GB.fit(train_features, train_label) ENS_model = ENSTest.fit(train_features_std, train_label) In?[31]: #為什么模型融合公式是這樣的？ Final_score = (np.exp(GB_model.predict(test_features)) + np.exp(ENS_model.predict(test_features_std))) / 2 In?[32]: #寫入csv文件 pd.DataFrame({'Id':test.Id, 'SalePrice':Final_score}).to_csv('submit.csv', index=False)

轉(zhuǎn)載于:https://www.cnblogs.com/RB26DETT/p/11566650.html

總結(jié)

以上是生活随笔為你收集整理的Kaggle: House Prices: Advanced Regression Techniques的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯，歡迎將生活随笔推薦給好友。