【飞桨学习赛：MarTech Challenge 点击反欺诈预测】第10名方案

---

该内容展示了一个模型从基线版本V1到V5的迭代过程，分数从86.746提升至89.0787。过程中进行了数据探索，处理缺失值和object类型字段，优化特征工程（如构造面积、时间差等特征），尝试LightGBM、XGBoost等模型，采用交叉验证，最终得到分数89.1093的最佳版本。

☞☞☞AI 智能聊天, 问答助手, AI 智能搜索, 免费无限量使用 DeepSeek R1 模型☜☜☜

In [ ]

#best版本-89.1093分数import pandas as pdimport warnings
warnings.filterwarnings('ignore')# 数据加载和去除Unnameed字段train = pd.read_csv('./train.csv')
test = pd.read_csv('./test1.csv')
train = train.iloc[:, 1:]
test = test.iloc[:,1:]
res = pd.DataFrame(test['sid'])# 去除数据探索发现问题的字段col = train.columns.tolist()
remove_list = ['lan', 'os','label', 'sid']for i in remove_list:
    col.remove(i)
features = train[col]
test_features = test[col]# 对osv进行数据清洗def osv_trans(x):
    x = str(x).replace('Android_', '').replace('Android ', '').replace('W', '')    if str(x).find('.')>0:
        temp_index1 = x.find('.')        if x.find(' ')>0:
            temp_index2 = x.find(' ')        else:
            temp_index2 = len(x) 
        if x.find('-')>0:
            temp_index2 = x.find('-')
            
        result = x[0:temp_index1] + '.' + x[temp_index1+1:temp_index2].replace('.', '')        try:            return float(result)        except:            print('有错误: '+x)            return 0
    try:        return float(x)    except:        print('有错误: '+x)        return 0features['osv'].fillna('8.1.0', inplace=True)
features['osv'] = features['osv'].apply(osv_trans)
test_features['osv'].fillna('8.1.0', inplace=True)
test_features['osv'] = test_features['osv'].apply(osv_trans)# 对timestamp进行数据清洗与特征变换,from datetime import datetime
features['timestamp'] = features['timestamp'].apply(lambda x: datetime.fromtimestamp(x/1000))
test_features['timestamp'] = test_features['timestamp'].apply(lambda x: datetime.fromtimestamp(x/1000))
temp = pd.DatetimeIndex(features['timestamp'])
features['year'] = temp.year
features['month'] = temp.month
features['day'] = temp.day
features['hour'] = temp.hour
features['minute'] = temp.minute
features['week_day'] = temp.weekday #星期几start_time = features['timestamp'].min()
features['time_diff'] = features['timestamp'] - start_time
features['time_diff'] = features['time_diff'].dt.days + features['time_diff'].dt.seconds/3600/24temp = pd.DatetimeIndex(test_features['timestamp'])
test_features['year'] = temp.year
test_features['month'] = temp.month
test_features['day'] = temp.day
test_features['hour'] = temp.hour
test_features['minute'] = temp.minute
test_features['week_day'] = temp.weekday #星期几 test_features['time_diff'] = test_features['timestamp'] - start_time
test_features['time_diff'] = test_features['time_diff'].dt.days + test_features['time_diff'].dt.seconds/3600/24features = features.drop(['timestamp'],axis = 1)
test_features = test_features.drop(['timestamp'],axis = 1)# 对version进行数据清洗与特征变换def version_trans(x):
    if x=='V3':        return 3
    if x=='v1':        return 1
    if x=='P_Final_6':        return 6
    if x=='V6':        return 6
    if x=='GA3':        return 3
    if x=='GA2':        return 2
    if x=='V2':        return 2
    if x=='50':        return 5
    return int(x)
features['version'] = features['version'].apply(version_trans)
test_features['version'] = test_features['version'].apply(version_trans)
features['version'] = features['version'].astype('int')
test_features['version'] = test_features['version'].astype('int')# 对lan进行数据清洗与特征变换 对于有缺失的lan 设置为22    lan_map = {'zh-CN': 1, 'zh_CN':2, 'Zh-CN': 3, 'zh-cn': 4, 'zh_CN_#Hans':5, 'zh': 6, 'ZH': 7, 'cn':8, 'CN':9, 'zh-HK': 10, 'tw': 11, 'TW': 12, 'zh-TW': 13,             'zh-MO':14, 'en':15, 'en-GB': 16, 'en-US': 17, 'ko': 18, 'ja': 19, 'it': 20, 'mi':21} 
train['lan'] = train['lan'].map(lan_map)
test['lan'] = test['lan'].map(lan_map)
train['lan'].fillna(22, inplace=True)
test['lan'].fillna(22, inplace=True)# 构造面积特征和构造相除特征features['dev_area'] = features['dev_height'] * features['dev_width']
test_features['dev_area'] = test_features['dev_height'] * test_features['dev_width']
features['dev_rato'] = features['dev_height'] / features['dev_width']
test_features['dev_rato'] = test_features['dev_height'] / test_features['dev_width']# APP版本与操作系统版本差features['version_osv'] = features['osv'] - features['version']
test_features['version_osv'] = test_features['osv'] - test_features['version']# 对fea_hash与fea1_hash特征变换features['fea_hash_len'] = features['fea_hash'].map(lambda x: len(str(x)))
features['fea1_hash_len'] = features['fea1_hash'].map(lambda x: len(str(x)))
features['fea_hash'] = features['fea_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
features['fea1_hash'] = features['fea1_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
test_features['fea_hash_len'] = test_features['fea_hash'].map(lambda x: len(str(x)))
test_features['fea1_hash_len'] = test_features['fea1_hash'].map(lambda x: len(str(x)))
test_features['fea_hash'] = test_features['fea_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
test_features['fea1_hash'] = test_features['fea1_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))#通过特征比，寻找关键特征，构造新特征，新特征字段 = 原始特征字段 + 1def find_key_feature(train, selected):
    temp = pd.DataFrame(columns = [0,1])
    temp0 = train[train['label'] == 0]
    temp1 = train[train['label'] == 1]
    temp[0] = temp0[selected].value_counts() / len(temp0) * 100
    temp[1] = temp1[selected].value_counts() / len(temp1) * 100
    temp[2] = temp[1] / temp[0]    #选出大于10倍的特征
    result = temp[temp[2] > 10].sort_values(2, ascending = False).index    return result
selected_cols = ['osv','apptype', 'carrier', 'dev_height', 'dev_ppi','dev_width', 'media_id', 
                 'package', 'version', 'fea_hash', 'location', 'fea1_hash','cus_type']
key_feature = {}for selected in selected_cols:
    key_feature[selected] = find_key_feature(train, selected)def f(x, selected):
    if x in key_feature[selected]:        return 1
    else:        return 0for selected in selected_cols:    if len(key_feature[selected]) > 0:
        features[selected+'1'] = features[selected].apply(f, args = (selected,))
        test_features[selected+'1'] = test_features[selected].apply(f, args = (selected,))        print(selected+'1 created')#CatBoost模型from catboost import CatBoostClassifierfrom sklearn.model_selection import StratifiedKFoldfrom sklearn.metrics import roc_auc_score
model=CatBoostClassifier(
            loss_function="Logloss",
            eval_metric="AUC",
            task_type="GPU",
            learning_rate=0.1,
            iterations=1000,
            random_seed=2025,
            od_type="Iter",
            depth=7)

n_folds =10 #十折交叉校验answers = []
mean_score = 0data_x=features
data_y=train['label']
sk = StratifiedKFold(n_splits=n_folds, shuffle=True, random_state=2025)
all_test = test_features.copy()for train, test in sk.split(data_x, data_y):  
    x_train = data_x.iloc[train]
    y_train = data_y.iloc[train]
    x_test = data_x.iloc[test]
    y_test = data_y.iloc[test]
    clf = model.fit(x_train,y_train, eval_set=(x_test,y_test),verbose=500) # 500条打印一条日志
    
    yy_pred_valid=clf.predict(x_test,prediction_type='Probability')[:,-1]    print('cat验证的auc:{}'.format(roc_auc_score(y_test, yy_pred_valid)))
    mean_score += roc_auc_score(y_test, yy_pred_valid) / n_folds
    
    y_pred_valid = clf.predict(all_test,prediction_type='Probability')[:,-1]
    answers.append(y_pred_valid) 
print('mean valAuc:{}'.format(mean_score))
cat_pre=sum(answers)/n_folds
cat_pre
res['label']=[1 if x>=0.5 else 0 for x in cat_pre]
res.to_csv('./baselinev6.csv',index=False)

       

有错误: f073b_changxiang_v01_b1b8_20180915
有错误: %E6%B1%9F%E7%81%B5OS+5.0
有错误: GIONEE_YNGA

       

项目思考的过程与baseline迭代版本

BaseLine V1_lgb--分数: 86.746

切换盘符：

jupyter notebook D:\

       

一、数据探索

1、去除Unnameed字段

train = train.iloc[:, 1:]
test = test.iloc[:,1:]

       

2、查看字段类型

写法1:

train.info()

       

写法2:

或者直接查看类型为object的列

train.select_dtypes(include='object').columns

       

发现以下字段为object类型需要进行数值变换

 7   lan         316720 non-null  object 
 10  os          500000 non-null  object 
 11  osv         493439 non-null  object 
 15  version     500000 non-null  object 
 16  fea_hash    500000 non-null  object

       

以lan为例查看里面数据情况

train['lan'].value_counts()

       

3、查看缺失值的个数

写法1：

train.isnull().sum()

       

写法2：

t = train.isnull().sum()
t[t>0]

       

发现以下字段缺少比较多

lan           183280osv             6561

       

4、唯一值的个数

查看唯一值的个数

features = train.columns.tolist()for feature in features:    if train[feature].nunique() ==1:        print(feature,train[feature].nunique())

       

发现os字段的唯一值个数太少

os 2

       

查看os

train['os'].value_counts()

       

发现os数据都为android

Openflow

一键极速绘图，赋能行业工作流

88 查看详情

android    303175Android    196825Name: os, dtype: int64

       

5、数据探索的结论

object类型字段有：lan、osv 、osv、version、fea_hash

缺失值较多的字段有：lan、osv

唯一值个数较少且意义不大：os

没有意义的字段：sid

BaselineV1中也先去除timestamp

6、特征的相关性分析(补充)

# 对特征列进行相关性分析import matplotlib.pyplot as plt
%matplotlib inlineimport seaborn as sns
plt.figure(figsize=(10,10))
sns.heatmap(train.corr(),cbar=True,annot=True,cmap='Blues')

       

二、数据预处理

最终去掉：【lan】【os】【osv】【version】【label】【sid】【timestamp】

remove_list = ['lan', 'os', 'osv', 'version', 'label', 'sid','timestamp']
col = features #字段名for i in remove_list:
    col.remove(i)
features = train[col]

       

三、特征工程

1、fea_hash特征变换

#查看数据值train['fea_hash'].value_counts()#查看统计信息train['fea_hash'].describe()#查看映射的长度特征情况train['fea_hash'].map(lambda x:len(str(x))).value_counts()

       

fea_hash进行特征变换

# fea_hash的长度为新特征features['fea_hash_len'] = features['fea_hash'].map(lambda x: len(str(x)))
features['fea1_hash_len'] = features['fea1_hash'].map(lambda x: len(str(x)))# 如果fea_hash很长，都归为0，否则为自己的本身features['fea_hash'] = features['fea_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
features['fea1_hash'] = features['fea1_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))

       

四、模型建立

test 做和train同样处理，利用lightgbm进行训练与预测，并保存，上诉过程全部合并代码如下：

#BaselineV1import pandas as pdimport warningsimport lightgbm as lgb
warnings.filterwarnings('ignore')# 数据加载train = pd.read_csv('./train.csv')
test = pd.read_csv('./test1.csv')# 去除Unnameed字段train = train.iloc[:, 1:]
test = test.iloc[:,1:]# 去除数据探索发现问题的字段col = train.columns.tolist()
remove_list = ['lan', 'os', 'osv', 'version', 'label', 'sid','timestamp']for i in remove_list:
    col.remove(i)
features = train[col]
test_features = test[col]# fea_hash特征变换features['fea_hash_len'] = features['fea_hash'].map(lambda x: len(str(x)))
features['fea1_hash_len'] = features['fea1_hash'].map(lambda x: len(str(x)))
features['fea_hash'] = features['fea_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
features['fea1_hash'] = features['fea1_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
test_features['fea_hash_len'] = test_features['fea_hash'].map(lambda x: len(str(x)))
test_features['fea1_hash_len'] = test_features['fea1_hash'].map(lambda x: len(str(x)))
test_features['fea_hash'] = test_features['fea_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
test_features['fea1_hash'] = test_features['fea1_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))#lightgbm进行训练与预测model = lgb.LGBMClassifier()
model.fit(features,train['label'])
result = model.predict(test_features)#res包括sid字段与label字段res = pd.DataFrame(test['sid'])
res['label'] = result#保存在csv中res.to_csv('./baselineV1.csv',index=False)

       

BaseLine V2_lgb--分数: 88.2007

一、特征工程优化

1、利用osv特征

# 对osv进行数据清洗def osv_trans(x):
    x = str(x).replace('Android_', '').replace('Android ', '').replace('W', '')    if str(x).find('.')>0:
        temp_index1 = x.find('.')        if x.find(' ')>0:
            temp_index2 = x.find(' ')        else:
            temp_index2 = len(x) 
        if x.find('-')>0:
            temp_index2 = x.find('-')
            
        result = x[0:temp_index1] + '.' + x[temp_index1+1:temp_index2].replace('.', '')        try:            return float(result)        except:            print('有错误: '+x)            return 0
    try:        return float(x)    except:        print('有错误: '+x)        return 0features['osv'].fillna('8.1.0', inplace=True)
features['osv'] = features['osv'].apply(osv_trans)
test_features['osv'].fillna('8.1.0', inplace=True)
test_features['osv'] = test_features['osv'].apply(osv_trans)

       

2、利用TimeStamp特征

提取时间多尺度并计算时间diff(时间差)

# 对timestamp进行数据清洗与特征变换from datetime import datetime
features['timestamp'] = features['timestamp'].apply(lambda x: datetime.fromtimestamp(x/1000))
test_features['timestamp'] = test_features['timestamp'].apply(lambda x: datetime.fromtimestamp(x/1000))
temp = pd.DatetimeIndex(features['timestamp'])
features['year'] = temp.year
features['month'] = temp.month
features['day'] = temp.day
features['hour'] = temp.hour
features['minute'] = temp.minute
features['week_day'] = temp.weekday #星期几start_time = features['timestamp'].min()
features['time_diff'] = features['timestamp'] - start_time
features['time_diff'] = features['time_diff'].dt.days + features['time_diff'].dt.seconds/3600/24temp = pd.DatetimeIndex(test_features['timestamp'])
test_features['year'] = temp.year
test_features['month'] = temp.month
test_features['day'] = temp.day
test_features['hour'] = temp.hour
test_features['minute'] = temp.minute
test_features['week_day'] = temp.weekday #星期几 test_features['time_diff'] = test_features['timestamp'] - start_time
test_features['time_diff'] = test_features['time_diff'].dt.days + test_features['time_diff'].dt.seconds/3600/24col = features.columns.tolist()
col.remove('timestamp')
features = features[col]
test_features = test_features[col]

       

3、利用Version特征

# 对version进行数据清洗与特征变换def version_trans(x):
    if x=='V3':        return 3
    if x=='v1':        return 1
    if x=='P_Final_6':        return 6
    if x=='V6':        return 6
    if x=='GA3':        return 3
    if x=='GA2':        return 2
    if x=='V2':        return 2
    if x=='50':        return 5
    return int(x)
features['version'] = features['version'].apply(version_trans)
test_features['version'] = test_features['version'].apply(version_trans)
features['version'] = features['version'].astype('int')
test_features['version'] = test_features['version'].astype('int')

       

二、模型建立

上诉过程合并代码如下：

import pandas as pdimport warningsimport lightgbm as lgb
warnings.filterwarnings('ignore')# 数据加载和去除Unnameed字段train = pd.read_csv('./train.csv')
test = pd.read_csv('./test1.csv')
train = train.iloc[:, 1:]
test = test.iloc[:,1:]# 去除数据探索发现问题的字段col = train.columns.tolist()
remove_list = ['lan', 'os','label', 'sid']for i in remove_list:
    col.remove(i)
features = train[col]
test_features = test[col]# 对osv进行数据清洗def osv_trans(x):
    x = str(x).replace('Android_', '').replace('Android ', '').replace('W', '')    if str(x).find('.')>0:
        temp_index1 = x.find('.')        if x.find(' ')>0:
            temp_index2 = x.find(' ')        else:
            temp_index2 = len(x) 
        if x.find('-')>0:
            temp_index2 = x.find('-')
            
        result = x[0:temp_index1] + '.' + x[temp_index1+1:temp_index2].replace('.', '')        try:            return float(result)        except:            print('有错误: '+x)            return 0
    try:        return float(x)    except:        print('有错误: '+x)        return 0features['osv'].fillna('8.1.0', inplace=True)
features['osv'] = features['osv'].apply(osv_trans)
test_features['osv'].fillna('8.1.0', inplace=True)
test_features['osv'] = test_features['osv'].apply(osv_trans)# 对timestamp进行数据清洗与特征变换,from datetime import datetime
features['timestamp'] = features['timestamp'].apply(lambda x: datetime.fromtimestamp(x/1000))
test_features['timestamp'] = test_features['timestamp'].apply(lambda x: datetime.fromtimestamp(x/1000))
temp = pd.DatetimeIndex(features['timestamp'])
features['year'] = temp.year
features['month'] = temp.month
features['day'] = temp.day
features['hour'] = temp.hour
features['minute'] = temp.minute
features['week_day'] = temp.weekday #星期几start_time = features['timestamp'].min()
features['time_diff'] = features['timestamp'] - start_time
features['time_diff'] = features['time_diff'].dt.days + features['time_diff'].dt.seconds/3600/24temp = pd.DatetimeIndex(test_features['timestamp'])
test_features['year'] = temp.year
test_features['month'] = temp.month
test_features['day'] = temp.day
test_features['hour'] = temp.hour
test_features['minute'] = temp.minute
test_features['week_day'] = temp.weekday #星期几 test_features['time_diff'] = test_features['timestamp'] - start_time
test_features['time_diff'] = test_features['time_diff'].dt.days + test_features['time_diff'].dt.seconds/3600/24features = features.drop(['timestamp'],axis = 1)
test_features = test_features.drop(['timestamp'],axis = 1)# 对version进行数据清洗与特征变换def version_trans(x):
    if x=='V3':        return 3
    if x=='v1':        return 1
    if x=='P_Final_6':        return 6
    if x=='V6':        return 6
    if x=='GA3':        return 3
    if x=='GA2':        return 2
    if x=='V2':        return 2
    if x=='50':        return 5
    return int(x)
features['version'] = features['version'].apply(version_trans)
test_features['version'] = test_features['version'].apply(version_trans)
features['version'] = features['version'].astype('int')
test_features['version'] = test_features['version'].astype('int')# 对fea_hash与fea1_hash特征变换features['fea_hash_len'] = features['fea_hash'].map(lambda x: len(str(x)))
features['fea1_hash_len'] = features['fea1_hash'].map(lambda x: len(str(x)))
features['fea_hash'] = features['fea_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
features['fea1_hash'] = features['fea1_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
test_features['fea_hash_len'] = test_features['fea_hash'].map(lambda x: len(str(x)))
test_features['fea1_hash_len'] = test_features['fea1_hash'].map(lambda x: len(str(x)))
test_features['fea_hash'] = test_features['fea_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))
test_features['fea1_hash'] = test_features['fea1_hash'].map(lambda x: 0 if len(str(x))>16 else int(x))#lightgbm进行训练与预测model = lgb.LGBMClassifier()
model.fit(features,train['label'])
result = model.predict(test_features)#res包括sid字段与label字段res = pd.DataFrame(test['sid'])
res['label'] = result#保存在csv中res.to_csv('./baselineV2.csv',index=False)print("已完成")

       

BaseLine V3_xgb--分数: 88.5073

一、特征工程优化

1、构造面积特征和相除特征

features['dev_area'] = features['dev_height'] * features['dev_width']
test_features['dev_area'] = test_features['dev_height'] * test_features['dev_width']
features['dev_rato'] = features['dev_height'] / features['dev_width']
test_features['dev_rato'] = test_features['dev_height'] / test_features['dev_width']

       

2、APP版本与操作系统版本差

features['version_osv'] = features['osv'] - features['version']
test_features['version_osv'] = test_features['osv'] - test_features['version']

       

二、xgboost模型

1、LightGBM 祖传参数

clf = lgb.LGBMClassifier(
            num_le*es=2**5-1, reg_alpha=0.25, reg_lambda=0.25, objective='multiclass',
            max_depth=-1, learning_rate=0.005, min_child_samples=3, random_state=2025,
            n_estimators=2000, subsample=1, colsample_bytree=1)
device = gpu
gpu_platform_id = 0gpu_device_id = 0

       

2、XGBoost祖传参数

model_xgb = xgb.XGBClassifier(
            max_depth=9, learning_rate=0.005, n_estimators=2000, 
            objective='multi:softprob', tree_method='gpu_hist', 
            subsample=0.8, colsample_bytree=0.8, 
            min_child_samples=3, eval_metric='logloss', reg_lambda=0.5)

       

3、使用xgboost并使用祖传参数

%%time#lightgbm进行训练与预测import xgboost as xgb
model_xgb = xgb.XGBClassifier(
            max_depth=15, learning_rate=0.05, n_estimators=5000, 
            objective='binary:logistic', tree_method='gpu_hist', 
            subsample=0.8, colsample_bytree=0.8, 
            min_child_samples=3, eval_metric='auc', reg_lambda=0.5
        )
model_xgb.fit(features,train['label'])
result_xgb = model.predict(test_features)
res = pd.DataFrame(test['sid'])
res['label'] = result_xgb
res.to_csv('./baselineV3.csv',index=False)print("已完成")

       

使用了xgboost的祖传参数

参数	含义
max_depth	含义：树的最大深度，用来避免过拟合的。max_depth越大，模型会学到更具体更局部的样本，需要使用CV函数来进行调优。  默认值：6，典型值：3-10。 调参：值越大，越容易过拟合；值越小，越容易欠拟合。
learning_rate	含义：学习率，控制每次迭代更新权重时的步长 默认值：0.3，典型值：0.01-0.2。  调参：值越小，训练越慢。
n_estimators	总共迭代的次数，即决策树的个数，相当于训练的轮数
objective	回归任务：reg:linear (默认) reg: logistic  二分类 binary:logistic (概率) binary：logitraw (类别)  多分类 multi：softmax num_class=n (返回类别) multi：softprob num_class=n(返回概率)
tree_method	可调用gpu：gpu_hist。使用功能的树的构建方法，hist代表使用直方图优化的近似贪婪的算法
subsample	含义：训练样本采样率（行采样），训练每棵树时，使用的数据占全部训练集的比例。这个参数控制对于每棵树，随机采样的比例。 减小这个参数的值，算法会更加保守，避免过拟合。但是，如果这个值设置得过小，它可能会导致欠拟合。 默认值：1，典型值：0.5-1。 调参：防止过拟合。
colsample_bytree	含义：训练每棵树时，使用的数据占全部训练集的比例。默认值为1，典型值为0.5-1。和GBM中的subsample参数一模一样。这个参数控制对于每棵树，随机采样的比例。 减小这个参数的值，算法会更加保守，避免过拟合。但是，如果这个值设置得过小，它可能会导致欠拟合。  典型值：0.5-1  调参：防止过拟合。
min_child_samples
eval_metric	用户可以添加多种评价指标，对于Python用户要以list传递参数对给程序 可供的选择如下:  回归任务(默认rmse) ：rmse--均方根误差 mae--平均绝对误差  分类任务(默认error) ： auc--roc曲线下面积 error--错误率（二分类） merror--错误率（多分类） logloss--负对数似然函数（二分类） mlogloss--负对数似然函数（多分类）
reg_lambda	L2正则化系数

4、可视化的方式查看特征的重要程度

from xgboost import plot_importanceimport matplotlib.pyplot as plt
plot_importance(model_xgb)

       

BaseLine V4_xgb--分数: 88.946

一、使用十折交叉验证优化

%%time# 定义10折子模型from sklearn.model_selection import StratifiedKFoldfrom sklearn.metrics import accuracy_scoredef xgb_model(clf,train_x,train_y,test):
    sk=StratifiedKFold(n_splits=10,random_state=2025,shuffle = True)
    prob=[]
    mean_acc=0
    for k,(train_index,val_index) in enumerate(sk.split(train_x,train_y)):
        train_x_real=train_x.iloc[train_index]
        train_y_real=train_y.iloc[train_index]
        val_x=train_x.iloc[val_index]
        val_y=train_y.iloc[val_index]        #模型训练及验证集测试
        clf=clf.fit(train_x_real,train_y_real)
        val_y_pred=clf.predict(val_x)
        acc_val=accuracy_score(val_y,val_y_pred)        print('第{}个子模型 accuracy{}'.format(k+1,acc_val))
        mean_acc+=mean_acc/10
        #预测测试集
        test_y_pred=clf.predict_proba(test)
        prob.append(test_y_pred)    print(mean_acc)
    mean_prob=sum(prob)/10
    return mean_prob 
 
import xgboost as xgb
model_xgb2 = xgb.XGBClassifier(
            max_depth=15, learning_rate=0.005, n_estimators=5300, 
            objective='binary:logistic', tree_method='gpu_hist', 
            subsample=0.7, colsample_bytree=0.7, 
            min_child_samples=3, eval_metric='auc', reg_lambda=0.5
        )
result_xgb=xgb_model(model_xgb2,features,train['label'],test_features) 
result_xgb2=[x[1] for x in result_xgb]
result_xgb2=[1 if x>=0.5 else 0 for x in result_xgb2]
 
res = pd.DataFrame(test['sid'])
res['label'] = result_xgb2
res.to_csv('./baselineV4.csv', index=False)print('已完成')

       

BaseLine V5_xgb--分数: 89.0787

一、特征工程优化

通过特征比，寻找关键特征，构造新特征，新特征字段 = 原始特征字段 + 1

#通过特征比，寻找关键特征，构造新特征，新特征字段 = 原始特征字段 + 1def find_key_feature(train, selected):
    temp = pd.DataFrame(columns = [0,1])
    temp0 = train[train['label'] == 0]
    temp1 = train[train['label'] == 1]
    temp[0] = temp0[selected].value_counts() / len(temp0) * 100
    temp[1] = temp1[selected].value_counts() / len(temp1) * 100
    temp[2] = temp[1] / temp[0]    #选出大于10倍的特征
    result = temp[temp[2] > 10].sort_values(2, ascending = False).index    return result

selected_cols = ['osv','apptype', 'carrier', 'dev_height', 'dev_ppi','dev_width', 'media_id', 
                 'package', 'version', 'fea_hash', 'location', 'fea1_hash','cus_type']
key_feature = {}for selected in selected_cols:
    key_feature[selected] = find_key_feature(train, selected)
key_featuredef f(x, selected):
    if x in key_feature[selected]:        return 1
    else:        return 0for selected in selected_cols:    if len(key_feature[selected]) > 0:
        features[selected+'1'] = features[selected].apply(f, args = (selected,))
        test_features[selected+'1'] = test_features[selected].apply(f, args = (selected,))        print(selected+'1 created')

   

以上就是【飞桨学习赛：MarTech Challenge 点击反欺诈预测】第10名方案的详细内容，更多请关注其它相关文章！

# git  # 操作系统  # ai  # 系统版本  # yy  # red  # python  # 加载  # 迭代  # 李老师网站运营推广  # 关键词排名优化公司  # 北京好的网站建设收费  # 营销推广方案ppt内容主题海报  # 前端seo关键词  # 怎样选择seo公司  # 谷歌seo优化时间  # 晋城抖音推广关键词排名  # 蓟县文教书籍网站建设  # 中文网  # 默认值  # 典型值  # 棵树  # 一言  # 新特征  # 有错误  # type  # fig  # igs  # 海外seo代理哪个好 

相关栏目： 【 Google疑问12 】 【 Facebook疑问10 】 【 优化推广96088 】 【 技术知识133117 】 【 IDC资讯59369 】 【 网络运营7196 】 【 IT资讯61894 】

相关推荐： 能抓取玻璃碎片、水下透明物，清华提出通用型透明物体抓取框架，成功率极高  码刻 | 48小时Hackathon，源码见证新生代AI创新的发生  华为推出两款商用 AI 大模型存储新品，支持 1200 万 IOPS 性能  看懂AI，找到增长新势能 | 笔记侠AI峰会等你来  《爱康未来之夜嘉宾官宣，携手共赴AI未来》  创新科学家成功研发FAST激光靶标维护机器人  苹果在韩举办首届中小企业智能制造论坛，加速推动工业4.0发展  无人机巡检方案是什么，该如何选择适合的巡检方案  讯飞听见会写“会议摘要”功能全面升级，AI更懂你的关注点  人工智能颠覆软件测试四大方式  GPT-4是如何工作的？哈佛教授亲自讲授  深度学习模型综述：用于3D MRI和CT扫描的应用  鸿蒙智能座舱的AI大模型革新，引领智能座舱领域的变革吗？  食品分销跨国企业Sysco CIDO：我们的增长秘诀是以IT为中心  华为云盘古大模型3.0发布 AI云服务同时上线：200亿亿次性能  创新全场景清洁方案！海尔商用机器人首发上市  Bing 聊天机器人现支持在桌面端用语音提问  城市在采用人工智能方面进展如何？  这款在《自然通讯》发表的机器人，为变形金刚来到现实创造可能性  OpenAI 静默关闭 AI 文本检测工具，准确率仅为 26%  Meta 人工智能业务落后竞争对手，研究人员大量离职成重要原因  Meta Connect 2025已确定时间为9月27-28，主题涵盖Quest 3与AI技术  OpenAI高管：AI能创造新的就业机会 但也会淘汰一些  「从未被制造出的最重要机器」，艾伦·图灵及图灵机那些事  如何提高集群协作效率？中外团队合作研发基于均值偏移的机器人队形控制策略  苹果AI战略与微软谷歌大相径庭，到底是领先还是落后？  人工智能快速发展 打开就业新空间  Unity 内测 Safe Voice 服务，利用 AI 自动识别玩家不当聊天内容  ChatGPT设计出的第一个机器人来了！【附人工智能行业预测】  京东 AI 大模型官宣 7 月 13 日发布，还有重磅合作  AI 大模型重塑软件开发，有哪些落地前景和痛点？| ArchSummit  禁止艺术家使用 AI 创作《龙与地下城》游戏插图的决定已在 D&D Beyond 生效  磐镭发布全新 GeForce RTX 4080 ARMOUR 显卡，售价为 9499 元  笔神作文声讨学而思AI大模型 称用“爬虫”技术盗取数据  十个AI算法常用库J*a版  视觉中国宣布推出AI灵感绘图、画面扩展功能  微软 Copilot 团队主管呼吁用户与 AI 交流时应使用恰当的礼貌用语  人工智能如何与智能家居集成  13条咒语挖掘GPT-4最大潜力，Github万星AI导师火了，网友：隔行再也不隔山了  元宇宙迈入2.0时代，它和生成式人工智能有何关联吗？  Meta发布语音AI模型 Voicebox 助虚拟助手与NPC对话  云深处与昇腾CANN携手合作：开设ROS四足机器狗开发训练营  ​布局智能物联新时代,中国移动“5G+物联网”亮相2025 MWC  2025VR&AR显示技术峰会展示歌尔光学最新一代光学模组  美踏控股推出创新人工智能大数据模型“心乐舞河”：虚拟人音舞社交的新体验  纪录片 《寻找人工智能》全集1080P超清  如何用AI开创智慧能源新时代？固德威正让能源“通人性”！  一图速览 | 十大脑机接口关键技术发布  搭载星火认知大模型 讯飞听见智慧屏开启AI办公新体验  跟着AI大热的“光模块”到底是什么？ 

 2025-08-01