ML之LiR&LassoR：利用boston房价数据集(PCA处理)采用线性回归和Lasso套索回归算法实现房价预测模型评估

一个处女座的程序猿 2020-11-25 23:29:07

2976

分类专栏： ML DataScience

最后发布:2020-11-25 23:29:07首次发布:2020-11-25 23:29:07

本文链接：https://blog.csdn.net/qq_41185868/article/details/110150258

版权

ML之LiR&LassoR：利用boston房价数据集(PCA处理)采用线性回归和Lasso套索回归算法实现房价预测模型评估

利用boston房价数据集(PCA处理)采用线性回归和Lasso套索回归算法实现房价预测模型评估

设计思路

输出结果

核心代码

利用boston房价数据集(PCA处理)采用线性回归和Lasso套索回归算法实现房价预测模型评估

设计思路

更新……

输出结果

   Id  MSSubClass MSZoning  ...  SaleType  SaleCondition SalePrice
0   1          60       RL  ...        WD         Normal    208500
1   2          20       RL  ...        WD         Normal    181500
2   3          60       RL  ...        WD         Normal    223500
3   4          70       RL  ...        WD        Abnorml    140000
4   5          60       RL  ...        WD         Normal    250000

[5 rows x 81 columns]
numeric_columns 36 ['LotFrontage', 'LotArea', 'OverallQual', 'OverallCond', 'YearBuilt', 'YearRemodAdd', 'MasVnrArea', 'BsmtFinSF1', 'BsmtFinSF2', 'BsmtUnfSF', 'TotalBsmtSF', '1stFlrSF', '2ndFlrSF', 'LowQualFinSF', 'GrLivArea', 'BsmtFullBath', 'BsmtHalfBath', 'FullBath', 'HalfBath', 'BedroomAbvGr', 'KitchenAbvGr', 'TotRmsAbvGrd', 'Fireplaces', 'GarageYrBlt', 'GarageCars', 'GarageArea', 'WoodDeckSF', 'OpenPorchSF', 'EnclosedPorch', '3SsnPorch', 'ScreenPorch', 'PoolArea', 'MiscVal', 'MoSold', 'YrSold', 'SalePrice']
(1460, 36)
   LotFrontage  LotArea  OverallQual  ...  MoSold  YrSold  SalePrice
0         65.0     8450            7  ...       2    2008     208500
1         80.0     9600            6  ...       5    2007     181500
2         68.0    11250            7  ...       9    2008     223500
3         60.0     9550            7  ...       2    2006     140000
4         84.0    14260            8  ...      12    2008     250000


依次统计每列缺失值元素个数： 
 36 [259, 0, 0, 0, 0, 0, 8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 81, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
Missing_data_Per_dict_0： (33, 0.9167, {'LotArea': 0.0, 'OverallQual': 0.0, 'OverallCond': 0.0, 'YearBuilt': 0.0, 'YearRemodAdd': 0.0, 'BsmtFinSF1': 0.0, 'BsmtFinSF2': 0.0, 'BsmtUnfSF': 0.0, 'TotalBsmtSF': 0.0, '1stFlrSF': 0.0, '2ndFlrSF': 0.0, 'LowQualFinSF': 0.0, 'GrLivArea': 0.0, 'BsmtFullBath': 0.0, 'BsmtHalfBath': 0.0, 'FullBath': 0.0, 'HalfBath': 0.0, 'BedroomAbvGr': 0.0, 'KitchenAbvGr': 0.0, 'TotRmsAbvGrd': 0.0, 'Fireplaces': 0.0, 'GarageCars': 0.0, 'GarageArea': 0.0, 'WoodDeckSF': 0.0, 'OpenPorchSF': 0.0, 'EnclosedPorch': 0.0, '3SsnPorch': 0.0, 'ScreenPorch': 0.0, 'PoolArea': 0.0, 'MiscVal': 0.0, 'MoSold': 0.0, 'YrSold': 0.0, 'SalePrice': 0.0})
Missing_data_Per_dict_Not0： (3, 0.0833, {'LotFrontage': 0.177397, 'MasVnrArea': 0.005479, 'GarageYrBlt': 0.055479})
Missing_data_Per_dict_under01： (2, 0.0556, {'MasVnrArea': 0.005479, 'GarageYrBlt': 0.055479})
依次计算每列缺失值元素占比： {'LotFrontage': 0.177397, 'MasVnrArea': 0.005479, 'GarageYrBlt': 0.055479}
data_Missing_dict {'LotFrontage': 0.1773972602739726, 'LotArea': 0.0, 'OverallQual': 0.0, 'OverallCond': 0.0, 'YearBuilt': 0.0, 'YearRemodAdd': 0.0, 'MasVnrArea': 0.005479452054794521, 'BsmtFinSF1': 0.0, 'BsmtFinSF2': 0.0, 'BsmtUnfSF': 0.0, 'TotalBsmtSF': 0.0, '1stFlrSF': 0.0, '2ndFlrSF': 0.0, 'LowQualFinSF': 0.0, 'GrLivArea': 0.0, 'BsmtFullBath': 0.0, 'BsmtHalfBath': 0.0, 'FullBath': 0.0, 'HalfBath': 0.0, 'BedroomAbvGr': 0.0, 'KitchenAbvGr': 0.0, 'TotRmsAbvGrd': 0.0, 'Fireplaces': 0.0, 'GarageYrBlt': 0.05547945205479452, 'GarageCars': 0.0, 'GarageArea': 0.0, 'WoodDeckSF': 0.0, 'OpenPorchSF': 0.0, 'EnclosedPorch': 0.0, '3SsnPorch': 0.0, 'ScreenPorch': 0.0, 'PoolArea': 0.0, 'MiscVal': 0.0, 'MoSold': 0.0, 'YrSold': 0.0, 'SalePrice': 0.0}
after dropna (1121, 36)
<class 'numpy.ndarray'>
      LotFrontage   LotArea  OverallQual  ...    MiscVal    MoSold    YrSold
0       -0.233570 -0.205885     0.570704  ...  -0.141407 -1.615345  0.153084
1        0.384834 -0.064358    -0.153825  ...  -0.141407 -0.498715 -0.596291
2       -0.109889  0.138702     0.570704  ...  -0.141407  0.990125  0.153084
3       -0.439705 -0.070512     0.570704  ...  -0.141407 -1.615345 -1.345665
4        0.549742  0.509132     1.295234  ...  -0.141407  2.106755  0.153084
...           ...       ...          ...  ...        ...       ...       ...
1116    -0.357251 -0.271480    -0.153825  ...  -0.141407  0.617915 -0.596291
1117     0.590968  0.375605    -0.153825  ...  -0.141407 -1.615345  1.651832
1118    -0.192343 -0.133030     0.570704  ...  14.947388 -0.498715  1.651832
1119    -0.109889 -0.049960    -0.878355  ...  -0.141407 -0.870925  1.651832
1120     0.178699 -0.022885    -0.878355  ...  -0.141407 -0.126505  0.153084

[1121 rows x 35 columns]
前10个主成分解释了数据中63.80%的变化
经过PCA后，进行第一层主成分分析-------------------------------------
[(0.16970682313415306, 'LotFrontage'), (0.1211669980146095, 'LotArea'), (0.3008665261375608, 'OverallQual'), (-0.1017783758120348, 'OverallCond'), (0.23754113423286216, 'YearBuilt'), (0.21067267847804322, 'YearRemodAdd'), (0.19125461510335365, 'MasVnrArea'), (0.14136511574315347, 'BsmtFinSF1'), (-0.013552848692716916, 'BsmtFinSF2'), (0.11439764110410199, 'BsmtUnfSF'), (0.259354275741638, 'TotalBsmtSF'), (0.2591780447881022, '1stFlrSF'), (0.11504305093601253, '2ndFlrSF'), (0.004231304806602964, 'LowQualFinSF'), (0.2877802164879641, 'GrLivArea'), (0.08317879411803167, 'BsmtFullBath'), (-0.02114280846249704, 'BsmtHalfBath'), (0.25499633884283257, 'FullBath'), (0.11080279874459822, 'HalfBath'), (0.1017767099777179, 'BedroomAbvGr'), (-0.01012145139988125, 'KitchenAbvGr'), (0.23572236584667458, 'TotRmsAbvGrd'), (0.17611466785004926, 'Fireplaces'), (0.23726651555979883, 'GarageYrBlt'), (0.2831568046802727, 'GarageCars'), (0.279827792756442, 'GarageArea'), (0.13036585867815073, 'WoodDeckSF'), (0.16664693092097654, 'OpenPorchSF'), (-0.08602539908222213, 'EnclosedPorch'), (0.010532579475601184, '3SsnPorch'), (0.02556170369869493, 'ScreenPorch'), (0.06246570190310543, 'PoolArea'), (-0.015493399959318557, 'MiscVal'), (0.028399126033275164, 'MoSold'), (-0.011129722622237775, 'YrSold')]
[(0.3008665261375608, 'OverallQual'), (0.2877802164879641, 'GrLivArea'), (0.2831568046802727, 'GarageCars'), (0.279827792756442, 'GarageArea'), (0.259354275741638, 'TotalBsmtSF'), (0.2591780447881022, '1stFlrSF'), (0.25499633884283257, 'FullBath'), (0.23754113423286216, 'YearBuilt'), (0.23726651555979883, 'GarageYrBlt'), (0.23572236584667458, 'TotRmsAbvGrd'), (0.21067267847804322, 'YearRemodAdd'), (0.19125461510335365, 'MasVnrArea'), (0.17611466785004926, 'Fireplaces'), (0.16970682313415306, 'LotFrontage'), (0.16664693092097654, 'OpenPorchSF'), (0.14136511574315347, 'BsmtFinSF1'), (0.13036585867815073, 'WoodDeckSF'), (0.1211669980146095, 'LotArea'), (0.11504305093601253, '2ndFlrSF'), (0.11439764110410199, 'BsmtUnfSF'), (0.11080279874459822, 'HalfBath'), (0.1017767099777179, 'BedroomAbvGr'), (0.08317879411803167, 'BsmtFullBath'), (0.06246570190310543, 'PoolArea'), (0.028399126033275164, 'MoSold'), (0.02556170369869493, 'ScreenPorch'), (0.010532579475601184, '3SsnPorch'), (0.004231304806602964, 'LowQualFinSF'), (-0.01012145139988125, 'KitchenAbvGr'), (-0.011129722622237775, 'YrSold'), (-0.013552848692716916, 'BsmtFinSF2'), (-0.015493399959318557, 'MiscVal'), (-0.02114280846249704, 'BsmtHalfBath'), (-0.08602539908222213, 'EnclosedPorch'), (-0.1017783758120348, 'OverallCond')]
经过PCA后，进行第二层主成分分析-------------------------------------
[(0.037140668512444255, 'LotFrontage'), (0.005762269875424171, 'LotArea'), (-0.02265545744738413, 'OverallQual'), (0.06797580738610676, 'OverallCond'), (-0.22034458100877843, 'YearBuilt'), (-0.11769773674122082, 'YearRemodAdd'), (-0.02330741979867707, 'MasVnrArea'), (-0.26830830083400875, 'BsmtFinSF1'), (-0.06776753790369254, 'BsmtFinSF2'), (0.10349973537774373, 'BsmtUnfSF'), (-0.2014230745261159, 'TotalBsmtSF'), (-0.14501101153644946, '1stFlrSF'), (0.43960496790131565, '2ndFlrSF'), (0.11932040000909688, 'LowQualFinSF'), (0.2706724094458561, 'GrLivArea'), (-0.2741406761479087, 'BsmtFullBath'), (-0.001880261013674545, 'BsmtHalfBath'), (0.12608264523927462, 'FullBath'), (0.23358978781221817, 'HalfBath'), (0.3864399252645517, 'BedroomAbvGr'), (0.12179545892853964, 'KitchenAbvGr'), (0.3371810668951179, 'TotRmsAbvGrd'), (0.06581774146310777, 'Fireplaces'), (-0.1834261688794573, 'GarageYrBlt'), (-0.04640661259007604, 'GarageCars'), (-0.08613653500685643, 'GarageArea'), (-0.047991361825782064, 'WoodDeckSF'), (0.03130768246434415, 'OpenPorchSF'), (0.13376424222015906, 'EnclosedPorch'), (-0.02564456693744644, '3SsnPorch'), (0.04211790221668751, 'ScreenPorch'), (0.03032238859229474, 'PoolArea'), (0.04968459727862472, 'MiscVal'), (0.02754218343139985, 'MoSold'), (-0.04555808126996797, 'YrSold')]
[(0.43960496790131565, '2ndFlrSF'), (0.3864399252645517, 'BedroomAbvGr'), (0.3371810668951179, 'TotRmsAbvGrd'), (0.2706724094458561, 'GrLivArea'), (0.23358978781221817, 'HalfBath'), (0.13376424222015906, 'EnclosedPorch'), (0.12608264523927462, 'FullBath'), (0.12179545892853964, 'KitchenAbvGr'), (0.11932040000909688, 'LowQualFinSF'), (0.10349973537774373, 'BsmtUnfSF'), (0.06797580738610676, 'OverallCond'), (0.06581774146310777, 'Fireplaces'), (0.04968459727862472, 'MiscVal'), (0.04211790221668751, 'ScreenPorch'), (0.037140668512444255, 'LotFrontage'), (0.03130768246434415, 'OpenPorchSF'), (0.03032238859229474, 'PoolArea'), (0.02754218343139985, 'MoSold'), (0.005762269875424171, 'LotArea'), (-0.001880261013674545, 'BsmtHalfBath'), (-0.02265545744738413, 'OverallQual'), (-0.02330741979867707, 'MasVnrArea'), (-0.02564456693744644, '3SsnPorch'), (-0.04555808126996797, 'YrSold'), (-0.04640661259007604, 'GarageCars'), (-0.047991361825782064, 'WoodDeckSF'), (-0.06776753790369254, 'BsmtFinSF2'), (-0.08613653500685643, 'GarageArea'), (-0.11769773674122082, 'YearRemodAdd'), (-0.14501101153644946, '1stFlrSF'), (-0.1834261688794573, 'GarageYrBlt'), (-0.2014230745261159, 'TotalBsmtSF'), (-0.22034458100877843, 'YearBuilt'), (-0.26830830083400875, 'BsmtFinSF1'), (-0.2741406761479087, 'BsmtFullBath')]
不进行PCA的线性回归的MSE是1644140595.6636596
前10个PCA主成分进行线性回归的MSE是1836601962.4751632
[1e-10, 1e-09, 1e-08, 1e-07, 1e-06, 1e-05, 0.0001, 0.001, 0.01, 0.1]
[1642818822.3530025, 1642818822.3529558, 1642818822.3524888, 1642818822.3471866, 1642818822.3005185, 1642818821.7415214, 1642818817.1179569, 1642818756.7038794, 1642818283.0732899, 1642813588.5752773]
[1e-10, 1e-09, 1e-08, 1e-07, 1e-06, 1e-05, 0.0001, 0.001, 0.01, 0.1]
[1836601962.4751682, 1836601962.4752123, 1836601962.475657, 1836601962.480097, 1836601962.5245085, 1836601962.9652405, 1836601967.4063494, 1836602011.8174434, 1836602455.9288514, 1836606882.1034737]

核心代码

PCA
class TruncatedSVD Found at: sklearn.decomposition._truncated_svd

class TruncatedSVD(TransformerMixin, BaseEstimator):
    """Dimensionality reduction using truncated SVD (aka LSA).
    
    This transformer performs linear dimensionality reduction by means of
    truncated singular value decomposition (SVD). Contrary to PCA, this
    estimator does not center the data before computing the singular value
    decomposition. This means it can work with sparse matrices
    efficiently.
    
    In particular, truncated SVD works on term count/tf-idf matrices as
    returned by the vectorizers in :mod:`sklearn.feature_extraction.text`. In
    that context, it is known as latent semantic analysis (LSA).
    
    This estimator supports two algorithms: a fast randomized SVD solver, 
     and
    a "naive" algorithm that uses ARPACK as an eigensolver on `X * X.T` or
    `X.T * X`, whichever is more efficient.
    

LinearRegression 
class LinearRegression Found at: sklearn.linear_model._base

class LinearRegression(MultiOutputMixin, RegressorMixin, LinearModel):
    """
    Ordinary least squares Linear Regression.
    
    LinearRegression fits a linear model with coefficients w = (w1, ..., wp)
    to minimize the residual sum of squares between the observed targets in
    the dataset, and the targets predicted by the linear approximation.
    

Lasso 
class Lasso Found at: sklearn.linear_model._coordinate_descent
class Lasso(ElasticNet):
    """Linear Model trained with L1 prior as regularizer (aka the Lasso)
    
    The optimization objective for Lasso is::
    
    (1 / (2 * n_samples)) * ||y - Xw||^2_2 + alpha * ||w||_1
    
    Technically the Lasso model is optimizing the same objective function as
    the Elastic Net with ``l1_ratio=1.0`` (no L2 penalty).
    
    Read more in the :ref:`User Guide <lasso>`.

一个处女座的程序猿

CSDN认证博客专家华为杯研电赛一等华为研数模一等奖国内外AI竞十

人工智能硕博生，目前兼职国内外多家头部人工智能公司的AI技术顾问。拥有十多项发明专利(6项)和软件著作权(9项)，多个国家级证书(2个国三级、3个国四级)，先后获得国内外“人工智能算法”竞赛(包括国家级、省市级等，一等奖5项、二等奖4项、三等奖2项)相关证书十多个，以上均以第一作者身份，并拥有省市校级个人荣誉证书十多项。正在撰写《人工智算法最新实战》一书，目前已37万字。