数据挖掘TASK4_建模调参

其他 2020-03-29 14:04:17 阅读次数: 0

建模与调参

学习目标
掌握机器学习模型的建模与调参过程

内容介绍

  1. 线性回归模型：

    • 线性回归对于特征的要求；

    • 处理长尾分布；

    • 理解线性回归模型；

  2. 模型性能验证：

    • 评价函数与目标函数；

    • 交叉验证方法；

    • 留一验证方法；

    • 针对时间序列问题的验证；

    • 绘制学习率曲线；

    • 绘制验证曲线；

  3. 嵌入式特征选择：

    • Lasso回归；

    • Ridge回归；

    • 决策树；

  4. 模型对比：

    • 常用线性模型；

    • 常用非线性模型；

  5. 模型调参：

    • 贪心调参方法；

    • 网格调参方法；

    • 贝叶斯调参方法；

代码示例


import pandas as pd

import numpy as np

import warnings

warnings.filterwarnings('ignore')


#定义reduce_men_usage函数，通过调整数据类型帮助我们减少数据所占内存空间

def reduce_mem_usage(df):

    start_men = df.memory_usage().sum()

    print('memory usage of dataframe is {:.2f} MB'.format(start_men))

    for col in df.columns:

        col_type = df[col].dtype

        if col_type != object:

            c_min = df[col].min()

            c_max = df[col].max()

            if str(col_type)[:3] == 'int':

                if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:

                    df[col] = df[col].astype(np.int8)

                elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:

                    df[col] = df[col].astype(np.int16)

                elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:

                    df[col] = df[col].astype(np.int32)

                elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:

                    df[col] = df[col].astype(np.int64)  

            else:

                if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:

                    df[col] = df[col].astype(np.float16)

                elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:

                    df[col] = df[col].astype(np.float32)

                else:

                    df[col] = df[col].astype(np.float64)

        else:

            df[col] = df[col].astype('category')

    end_men = df.memory_usage().sum()

    print('Memory usage after optimization is :{:.2f} MB'.format(end_men))

    print('Decreased by {:.1f}%'.format(100*(start_men - end_men)/start_men))

    return df



sample_feature = reduce_mem_usage(pd.read_csv('data_for_tree.csv'))

#sample_feature.head()

memory usage of dataframe is 62099624.00 MB

Memory usage after optimization is :16520255.00 MB

Decreased by 73.4%


continuous_feature_names = [x for x in sample_feature.columns if x not in ['price','brand','model','brand']]

print(continuous_feature_names)

sample_feature = sample_feature.dropna().replace('-', 0).reset_index(drop=True)

sample_feature['notRepairedDamage'] = sample_feature['notRepairedDamage'].astype(np.float32)

train = sample_feature[continuous_feature_names + ['price']]

train_x = train[continuous_feature_names]

train_y = train['price']

['SaleID', 'name', 'bodyType', 'fuelType', 'gearbox', 'power', 'kilometer', 'notRepairedDamage', 'seller', 'offerType', 'v_0', 'v_1', 'v_2', 'v_3', 'v_4', 'v_5', 'v_6', 'v_7', 'v_8', 'v_9', 'v_10', 'v_11', 'v_12', 'v_13', 'v_14', 'train', 'used_time', 'city', 'brand_amount', 'brand_price_average', 'brand_price_max', 'brand_price_median', 'brand_price_min', 'brand_price_std', 'brand_price_sum', 'power_bin']


#train_x.head()


train_y.head()

0    1850.0

1    6222.0

2    5200.0

3    8000.0

4    3500.0

Name: price, dtype: float64


#1、简单建模,训练线性回归模型，查看截距与权重

from sklearn.linear_model import LinearRegression

model = LinearRegression(normalize=True)

model = model.fit(train_x, train_y)

sorted(dict(zip(continuous_feature_names, model.coef_)).items(), key=lambda x:x[1], reverse=True)

[('v_6', 3367064.341641952),

 ('v_8', 700675.5609398864),

 ('v_9', 170630.27723221222),

 ('v_7', 32322.661932025392),

 ('v_12', 20473.670796989394),

 ('v_3', 17868.07954151005),

 ('v_11', 11474.938996718518),

 ('v_13', 11261.764560017724),

 ('v_10', 2683.920090609242),

 ('gearbox', 881.8225039249613),

 ('fuelType', 363.90425072161565),

 ('bodyType', 189.60271012074494),

 ('city', 44.9497512052328),

 ('power', 28.55390161675131),

 ('brand_price_median', 0.5103728134078974),

 ('brand_price_std', 0.45036347092632434),

 ('brand_amount', 0.1488112039506708),

 ('brand_price_max', 0.0031910186703149753),

 ('SaleID', 5.3559899198567324e-05),

 ('seller', 2.4531036615371704e-06),

 ('train', 4.246830940246582e-07),

 ('offerType', -7.235445082187653e-06),

 ('brand_price_sum', -2.175006868187898e-05),

 ('name', -0.00029800127130847845),

 ('used_time', -0.0025158943328449923),

 ('brand_price_average', -0.40490484510113794),

 ('brand_price_min', -2.246775348688707),

 ('power_bin', -34.42064411726649),

 ('v_14', -274.7841180776088),

 ('kilometer', -372.897526660709),

 ('notRepairedDamage', -495.19038446298714),

 ('v_0', -2045.0549573540754),

 ('v_5', -11022.986240523212),

 ('v_4', -15121.731109858125),

 ('v_2', -26098.29992055678),

 ('v_1', -45556.189297264835)]


from matplotlib import pyplot as plt

subsample_index = np.random.randint(low=0, high=len(train_y),size=50)#随机抽取50个点验证

plt.scatter(train_x['v_9'][subsample_index], train_y[subsample_index], color='black')

plt.scatter(train_x['v_9'][subsample_index], model.predict(train_x.loc[subsample_index]), color='blue')

plt.xlabel('v_9')

plt.ylabel('price')

plt.legend(['True Price','Predicted Price'],loc='upper right')

print('The predicted price is obvious different from true price')

plt.show()



The predicted price is obvious different from true price
![在这里插入图片描述](https://img-blog.csdnimg.cn/20200329134510981.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80Mzk1OTI0OA==,size_16,color_FFFFFF,t_70)






<Figure size 640x480 with 1 Axes>


import seaborn as sns

print('It is clear to see the price shows a typical exponential distribution')

plt.figure(figsize=(15,5))

plt.subplot(1,2,1)

sns.distplot(train_y)

plt.subplot(1,2,2)

sns.distplot(train_y[train_y < train_y.quantile(0.9)]) #将长尾截断

It is clear to see the price shows a typical exponential distribution


<matplotlib.axes._subplots.AxesSubplot at 0x17cbe231c18>

[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-qR3hcGGy-1585460408912)(output_7_3.png)]


#对标签进行log（x+1）变换使其贴近于正态分布，加一是为了防止底数为0

train_y_ln = np.log(train_y+1)

import seaborn as sns

print('The transformed price seems like normal distribution')

plt.figure(figsize=(15,5))

plt.subplot(1,2,1)

sns.distplot(train_y_ln)

plt.subplot(1,2,2)

sns.distplot(train_y_ln[train_y_ln < train_y.quantile(0.9)])

The transformed price seems like normal distribution

<matplotlib.axes._subplots.AxesSubplot at 0x17cbe53ffd0>

[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-4cBRUryw-1585460408915)(output_8_3.png)]


model = model.fit(train_x, train_y_ln)

print('intercept:'+str(model.intercept_))

sorted(dict(zip(continuous_feature_names, model.coef_)).items(), key=lambda x:x[1], reverse=True)



intercept:18.75074946557562

[('v_9', 8.052409900567515),

 ('v_5', 5.7642365966517515),

 ('v_12', 1.6182081236790782),

 ('v_1', 1.479831058296809),

 ('v_11', 1.1669016563609707),

 ('v_13', 0.9404711296034489),

 ('v_7', 0.713727308356328),

 ('v_3', 0.6837875771083226),

 ('v_0', 0.008500518010020237),

 ('power_bin', 0.00849796930289155),

 ('gearbox', 0.00792237727832305),

 ('fuelType', 0.006684769706828705),

 ('bodyType', 0.004523520092702963),

 ('power', 0.0007161894205359341),

 ('brand_price_min', 3.334351114747353e-05),

 ('brand_amount', 2.8978797042768103e-06),

 ('brand_price_median', 1.2571172872977267e-06),

 ('brand_price_std', 6.65917636342063e-07),

 ('brand_price_max', 6.194956307515807e-07),

 ('brand_price_average', 5.999345965093302e-07),

 ('SaleID', 2.1194170039646528e-08),

 ('seller', 9.978862181014847e-11),

 ('train', 7.958078640513122e-13),

 ('brand_price_sum', -1.5126504215909907e-10),

 ('offerType', -2.547437816247111e-10),

 ('name', -7.01551258888878e-08),

 ('used_time', -4.122479372354066e-06),

 ('city', -0.002218782481042724),

 ('v_14', -0.004234223418128389),

 ('kilometer', -0.013835866226882864),

 ('notRepairedDamage', -0.27027942349845646),

 ('v_4', -0.8315701200995309),

 ('v_2', -0.9470842241621843),

 ('v_10', -1.6261466689779176),

 ('v_8', -40.34300748761737),

 ('v_6', -238.79036385507334)]


plt.scatter(train_x['v_9'][subsample_index], train_y[subsample_index], color='black')

plt.scatter(train_x['v_9'][subsample_index], np.exp(model.predict(train_x.loc[subsample_index])), color='blue')

plt.xlabel('v_9')

plt.ylabel('price')

plt.legend(['True Price','Predicted Price'],loc='upper right')

print('The predicted price seems normal after np.log transforming')

plt.show()

The predicted price seems normal after np.log transforming

在这里插入图片描述


#2、五折交叉验证

#训练集，评估集，测试集。拿出训练集的一部分出来作为评估集，来对训练集生成的参数进行测试

from sklearn.model_selection import cross_val_score

from sklearn.metrics import mean_absolute_error,  make_scorer


def log_transfer(func):

    def wrapper(y, yhat):

        result = func(np.log(y),np.nan_to_num(np.log(yhat)))

        return result

    return wrapper



scores = cross_val_score(model, X=train_x, y=train_y, verbose=1, cv=5, scoring=make_scorer(log_transfer(mean_absolute_error)))

[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.7s finished


print('AVG:',np.mean(scores))

AVG: 1.3658023920313513


scores = pd.DataFrame(scores.reshape(1,-1))

scores.columns = ['cv' + str(x) for x in range(1, 6)]

scores.index = ['MAE']

scores

<tr style="text-align: right;">

  <th></th>

  <th>cv1</th>

  <th>cv2</th>

  <th>cv3</th>

  <th>cv4</th>

  <th>cv5</th>

</tr>

<tr>

  <th>MAE</th>

  <td>1.348304</td>

  <td>1.36349</td>

  <td>1.380712</td>

  <td>1.378401</td>

  <td>1.358105</td>

</tr>


import numpy as np

np.reshape(scores, [1,-1])

scores.columns = ['cv' + str(x) for x in range(1, 6)]

scores.index = ['MAE']

scores

<tr style="text-align: right;">

  <th></th>

  <th>cv1</th>

  <th>cv2</th>

  <th>cv3</th>

  <th>cv4</th>

  <th>cv5</th>

</tr>

<tr>

  <th>MAE</th>

  <td>1.348304</td>

  <td>1.36349</td>

  <td>1.380712</td>

  <td>1.378401</td>

  <td>1.358105</td>

</tr>


#3、模拟真实业务

#采用时间顺序对数据集进行分割，选靠前时间的4/5作为训练集，靠后的1/5作为验证集

sample_feature = sample_feature.reset_index(drop=True)

split_point = len(sample_feature)//5*4

train = sample_feature[:split_point].dropna()

val = sample_feature[split_point:].dropna()

train_x = train[continuous_feature_names]

train_y_ln = np.log(train['price'] + 1)

val_x = val[continuous_feature_names]

val_y_ln = np.log(val['price'] + 1)

model = model.fit(train_x, train_y_ln)

print('intercept:'+str(model.intercept_))

mean_absolute_error(val_y_ln, model.predict(val_x))

intercept:17.26478651939934

0.1957766416421094


#4、绘制学习曲线

from sklearn.model_selection import learning_curve, validation_curve

def plot_learning_curve(estimator, title, X, y, ylim=None, cv=None,n_jobs=1, train_size=np.linspace(.1, 1.0, 5 )):  

    plt.figure()  

    plt.title(title)  

    if ylim is not None:  

        plt.ylim(*ylim)  

    plt.xlabel('Training example')  

    plt.ylabel('score')  

    train_sizes, train_scores, test_scores = learning_curve(estimator, X, y, cv=cv, n_jobs=n_jobs, train_sizes=train_size, scoring = make_scorer(mean_absolute_error))  

    train_scores_mean = np.mean(train_scores, axis=1)  

    train_scores_std = np.std(train_scores, axis=1)  

    test_scores_mean = np.mean(test_scores, axis=1)  

    test_scores_std = np.std(test_scores, axis=1)  

    plt.grid()#区域  

    plt.fill_between(train_sizes, train_scores_mean - train_scores_std,  

                     train_scores_mean + train_scores_std, alpha=0.1,  

                     color="r")  

    plt.fill_between(train_sizes, test_scores_mean - test_scores_std,  

                     test_scores_mean + test_scores_std, alpha=0.1,  

                     color="g")  

    plt.plot(train_sizes, train_scores_mean, 'o-', color='r',  

             label="Training score")  

    plt.plot(train_sizes, test_scores_mean,'o-',color="g",  

             label="Cross-validation score")  

    plt.legend(loc="best")  

    return plt 

plot_learning_curve(LinearRegression(), 'Liner_model', train_x[:1000], train_y_ln[:1000], ylim=(0.0, 0.5), cv=5, n_jobs=1) 

<module 'matplotlib.pyplot' from 'C:\\Users\\lenovo\\Anaconda3\\lib\\site-packages\\matplotlib\\pyplot.py'>

在这里插入图片描述


#模型调参

#1、通过线性回归，加入两种正则化方法，变成岭回归和Lasso回归

from sklearn.linear_model import LinearRegression

from sklearn.linear_model import Ridge

from sklearn.linear_model import Lasso

train = sample_feature[continuous_feature_names + ['price']].dropna()

train_X = train[continuous_feature_names]

train_y = train['price']

train_y_ln = np.log(train_y + 1)

#三种模型

result = {}

models = [LinearRegression(), Ridge(), Lasso()]

for model in models:

    model_name = str(model).split('(')[0]

    scores = cross_val_score(model, X=train_X, y=train_y_ln, verbose=0, cv=5, scoring=make_scorer(mean_absolute_error))

    result[model_name] = scores

    print(model_name+' is finished')

result = pd.DataFrame(result)

result.index = ['cv' + str(x) for x in range(1, 6)]

result

LinearRegression is finished

Ridge is finished

Lasso is finished

<tr style="text-align: right;">

  <th></th>

  <th>LinearRegression</th>

  <th>Ridge</th>

  <th>Lasso</th>

</tr>

<tr>

  <th>cv1</th>

  <td>0.190792</td>

  <td>0.194832</td>

  <td>0.383899</td>

</tr>

<tr>

  <th>cv2</th>

  <td>0.193758</td>

  <td>0.197632</td>

  <td>0.381894</td>

</tr>

<tr>

  <th>cv3</th>

  <td>0.194132</td>

  <td>0.198123</td>

  <td>0.384090</td>

</tr>

<tr>

  <th>cv4</th>

  <td>0.191825</td>

  <td>0.195670</td>

  <td>0.380526</td>

</tr>

<tr>

  <th>cv5</th>

  <td>0.195758</td>

  <td>0.199676</td>

  <td>0.383611</td>

</tr>


#线性回归

model = LinearRegression().fit(train_X, train_y_ln)

print('intercept:'+ str(model.intercept_))

sns.barplot(abs(model.coef_), continuous_feature_names)

intercept:18.750750028424832

在这里插入图片描述


#岭回归

model = Ridge().fit(train_X, train_y_ln)

print('intercept:'+ str(model.intercept_))

sns.barplot(abs(model.coef_), continuous_feature_names)

intercept:4.671709788130855

<matplotlib.axes._subplots.AxesSubplot at 0x17cbb77c6a0>

在这里插入图片描述


#LASSO回归

model = Lasso().fit(train_X, train_y_ln)

print('intercept:'+ str(model.intercept_))

sns.barplot(abs(model.coef_), continuous_feature_names)

intercept:8.67218477988307

<matplotlib.axes._subplots.AxesSubplot at 0x17cbb8be240>

在这里插入图片描述


#2、非线性回归

from sklearn.linear_model import LinearRegression

from sklearn.svm import SVC

from sklearn.tree import DecisionTreeRegressor

from sklearn.ensemble import RandomForestRegressor

from sklearn.ensemble import GradientBoostingRegressor

from sklearn.neural_network import MLPRegressor

from xgboost.sklearn import XGBRegressor

from lightgbm.sklearn import LGBMRegressor


models = [LinearRegression(),

          DecisionTreeRegressor(),

          RandomForestRegressor(),

          GradientBoostingRegressor(),

          MLPRegressor(solver='lbfgs', max_iter=100), 

          XGBRegressor(n_estimators = 100, objective='reg:squarederror'), 

          LGBMRegressor(n_estimators = 100)]


result = dict()

for model in models:

    model_name = str(model).split('(')[0]

    scores = cross_val_score(model, X=train_X, y=train_y_ln, verbose=0, cv = 5, scoring=make_scorer(mean_absolute_error))

    result[model_name] = scores

    print(model_name + ' is finished')

LinearRegression is finished

DecisionTreeRegressor is finished

RandomForestRegressor is finished

GradientBoostingRegressor is finished

MLPRegressor is finished

XGBRegressor is finished

LGBMRegressor is finished


result = pd.DataFrame(result)

result.index = ['cv' + str(x) for x in range(1, 6)]

result

<tr style="text-align: right;">

  <th></th>

  <th>LinearRegression</th>

  <th>DecisionTreeRegressor</th>

  <th>RandomForestRegressor</th>

  <th>GradientBoostingRegressor</th>

  <th>MLPRegressor</th>

  <th>XGBRegressor</th>

  <th>LGBMRegressor</th>

</tr>

<tr>

  <th>cv1</th>

  <td>0.190792</td>

  <td>0.198679</td>

  <td>0.140822</td>

  <td>0.168900</td>

  <td>285.562549</td>

  <td>0.142367</td>

  <td>0.141542</td>

</tr>

<tr>

  <th>cv2</th>

  <td>0.193758</td>

  <td>0.193387</td>

  <td>0.143273</td>

  <td>0.171831</td>

  <td>572.989841</td>

  <td>0.140923</td>

  <td>0.145501</td>

</tr>

<tr>

  <th>cv3</th>

  <td>0.194132</td>

  <td>0.189258</td>

  <td>0.142621</td>

  <td>0.170875</td>

  <td>300.496953</td>

  <td>0.139393</td>

  <td>0.143887</td>

</tr>

<tr>

  <th>cv4</th>

  <td>0.191825</td>

  <td>0.190014</td>

  <td>0.142087</td>

  <td>0.169064</td>

  <td>2114.730472</td>

  <td>0.137492</td>

  <td>0.142497</td>

</tr>

<tr>

  <th>cv5</th>

  <td>0.195758</td>

  <td>0.204785</td>

  <td>0.144554</td>

  <td>0.174094</td>

  <td>353.180810</td>

  <td>0.143732</td>

  <td>0.144852</td>

</tr>


#模型调参

objective = ['regression', 'regression_l1', 'mape', 'huber', 'fair']

num_leaves = [3,5,10,15,20,40, 55]

max_depth = [3,5,10,15,20,40, 55]

bagging_fraction = []

feature_fraction = []

drop_rate = []

#1、贪心算法

best_obj = dict()

for obj in objective:

    model = LGBMRegressor(objective=obj)

    score = np.mean(cross_val_score(model, X=train_X, y=train_y_ln, verbose=0, cv = 5, scoring=make_scorer(mean_absolute_error)))

    best_obj[obj] = score

    

best_leaves = dict()

for leaves in num_leaves:

    model = LGBMRegressor(objective=min(best_obj.items(), key=lambda x:x[1])[0], num_leaves=leaves)

    score = np.mean(cross_val_score(model, X=train_X, y=train_y_ln, verbose=0, cv = 5, scoring=make_scorer(mean_absolute_error)))

    best_leaves[leaves] = score

    

best_depth = dict()

for depth in max_depth:

    model = LGBMRegressor(objective=min(best_obj.items(), key=lambda x:x[1])[0],

                          num_leaves=min(best_leaves.items(), key=lambda x:x[1])[0],

                          max_depth=depth)

    score = np.mean(cross_val_score(model, X=train_X, y=train_y_ln, verbose=0, cv = 5, scoring=make_scorer(mean_absolute_error)))

    best_depth[depth] = score

sns.barplot(x=['0_initial','1_turning_obj','2_turning_leaves','3_turning_depth'], y=[0.143 ,min(best_obj.values()), min(best_leaves.values()), min(best_depth.values())])

<matplotlib.axes._subplots.AxesSubplot at 0x17cbcd0f048>

在这里插入图片描述


#grid-search调参(穷举搜索)

from sklearn.model_selection import GridSearchCV


parameters = {'objective': objective , 'num_leaves': num_leaves, 'max_depth': max_depth}

model = LGBMRegressor()

clf = GridSearchCV(model, parameters, cv=5)

clf = clf.fit(train_X, train_y)

clf.best_params_

{'max_depth': 15, 'num_leaves': 55, 'objective': 'regression'}


model = LGBMRegressor(objective='regression',

                          num_leaves=55,

                          max_depth=15)


np.mean(cross_val_score(model, X=train_X, y=train_y_ln, verbose=0, cv = 5, scoring=make_scorer(mean_absolute_error)))

0.13754980533444577


#贝叶斯调参

from bayes_opt import BayesianOptimization


def rf_cv(num_leaves, max_depth, subsample, min_child_samples):

    val = cross_val_score(

        LGBMRegressor(objective = 'regression_l1',

            num_leaves=int(num_leaves),

            max_depth=int(max_depth),

            subsample = subsample,

            min_child_samples = int(min_child_samples)

        ),

        X=train_X, y=train_y_ln, verbose=0, cv = 5, scoring=make_scorer(mean_absolute_error)

    ).mean()

    return 1 - val


rf_bo = BayesianOptimization(

    rf_cv,

    {

    'num_leaves': (2, 100),

    'max_depth': (2, 100),

    'subsample': (0.1, 1),

    'min_child_samples' : (2, 100)

    }

)


rf_bo.maximize()

|   iter    |  target   | max_depth | min_ch... | num_le... | subsample |

-------------------------------------------------------------------------

|  1        |  0.8625   |  98.27    |  16.21    |  46.74    |  0.5154   |

|  2        |  0.867    |  60.17    |  24.19    |  73.85    |  0.5303   |

|  3        |  0.8678   |  20.73    |  49.05    |  79.91    |  0.9991   |

|  4        |  0.8686   |  11.38    |  33.55    |  96.73    |  0.106    |

|  5        |  0.8583   |  28.24    |  88.14    |  32.07    |  0.54     |

|  6        |  0.8692   |  99.18    |  99.2     |  99.89    |  0.5816   |

|  7        |  0.8692   |  98.37    |  3.355    |  98.11    |  0.3583   |

|  8        |  0.8505   |  5.726    |  3.353    |  99.91    |  0.9506   |

|  9        |  0.8398   |  4.988    |  98.7     |  95.51    |  0.2637   |

|  10       |  0.802    |  98.82    |  96.37    |  3.977    |  0.7117   |

|  11       |  0.7719   |  6.261    |  12.23    |  2.926    |  0.9965   |

|  12       |  0.8668   |  56.81    |  23.78    |  71.71    |  0.1635   |

|  13       |  0.8684   |  99.3     |  46.5     |  86.75    |  0.1027   |

|  14       |  0.8693   |  51.32    |  77.08    |  99.54    |  0.1632   |

|  15       |  0.8678   |  17.64    |  47.26    |  78.37    |  0.5125   |

|  16       |  0.8654   |  67.56    |  99.3     |  62.61    |  0.1608   |

|  17       |  0.8694   |  48.5     |  43.38    |  99.52    |  0.1868   |

|  18       |  0.8632   |  57.29    |  61.38    |  49.45    |  0.2046   |

|  19       |  0.8666   |  95.77    |  3.698    |  71.83    |  0.5748   |

|  20       |  0.8689   |  85.61    |  76.58    |  98.76    |  0.6544   |

|  21       |  0.8692   |  70.03    |  98.23    |  99.73    |  0.3661   |

|  22       |  0.8692   |  97.84    |  27.73    |  99.84    |  0.212    |

|  23       |  0.8678   |  53.85    |  61.55    |  80.23    |  0.1136   |

|  24       |  0.8691   |  51.15    |  4.563    |  99.3     |  0.1271   |

|  25       |  0.8694   |  35.05    |  25.77    |  99.39    |  0.8209   |

|  26       |  0.8692   |  72.39    |  20.52    |  99.82    |  0.3174   |

|  27       |  0.8691   |  99.66    |  71.71    |  99.5     |  0.2219   |

|  28       |  0.8693   |  25.56    |  42.44    |  99.05    |  0.1066   |

|  29       |  0.8664   |  33.56    |  81.97    |  69.32    |  0.2377   |

|  30       |  0.868    |  87.97    |  93.77    |  85.64    |  0.1569   |

=========================================================================


1 - rf_bo.max['target']

0.1305975267548991


plt.figure(figsize=(13,5))

sns.barplot(x=['0_origin','1_log_transfer','2_L1_&_L2','3_change_model','4_parameter_turning'], y=[1.36 ,0.19, 0.19, 0.14, 0.13])

<matplotlib.axes._subplots.AxesSubplot at 0x17cbccdf198>

在这里插入图片描述

北海星
发布了6 篇原创文章 · 获赞 0 · 访问量 92
私信 关注

猜你喜欢

转载自blog.csdn.net/weixin_43959248/article/details/105157336
今日推荐
周排行
成为C++高手之宏与枚举
在CAD二次开发中使用进度条
Js插件ECharts，HighCharts学习网址整理
Celery提交任务出错(on windows.)
cephfs内核客户端性能追踪
thinkphp中PHPExcel用法
EntityFramework动态组合多排序字段
汇编语言（八）实验9 根据材料编程
安装ubuntu后必须做的事情（对我而言）
JS函数式编程
每日归档
更多
2024-10-22(0)
2024-10-21(0)
2024-10-20(0)
2024-10-19(0)
2024-10-18(0)
2024-10-17(0)
2024-10-16(0)
2024-10-15(0)
2024-10-14(0)
2024-10-13(0)

代码天地 © 2018 codetd.com

境外服务器   最新电视剧2022