天池推荐系统排序模型+模型融合

排序模型

通过召回几经根据数据的不同特征缩减了问题的规模，接下来就是使用监督学习来对各种特征使用训练集进行表征，然后对测试集进行预测，得到测试集中的每个候选集用户点击的概率作为最终的结果。

排序阶段选择了三个比较有代表性的排序模型，它们分别是：

1. LGB的排序模型
2. LGB的分类模型
3. 深度学习的分类模型DIN

得到了最终的排序模型输出的结果之后，还选择了两种比较经典的模型集成的方法：

1. 输出结果加权融合
2. Staking（将模型的输出结果再使用一个简单模型进行预测）

import numpy as np
import pandas as pd
import pickle
from tqdm import tqdm
import gc, os
import time
from datetime import datetime
import lightgbm as lgb
from sklearn.preprocessing import MinMaxScaler
import warnings
warnings.filterwarnings('ignore')

读取排序特征

data_path = './data_raw/'
save_path = './temp_results/'
offline = False

# 重新读取数据的时候，发现click_article_id是一个浮点数，所以将其转换成int类型
trn_user_item_feats_df = pd.read_csv(save_path + 'trn_user_item_feats_df.csv')
trn_user_item_feats_df['click_article_id'] = trn_user_item_feats_df['click_article_id'].astype(int)

if offline:
    val_user_item_feats_df = pd.read_csv(save_path + 'val_user_item_feats_df.csv')
    val_user_item_feats_df['click_article_id'] = val_user_item_feats_df['click_article_id'].astype(int)
else:
    val_user_item_feats_df = None
    
tst_user_item_feats_df = pd.read_csv(save_path + 'tst_user_item_feats_df.csv')
tst_user_item_feats_df['click_article_id'] = tst_user_item_feats_df['click_article_id'].astype(int)

# 做特征的时候为了方便，给测试集也打上了一个无效的标签，这里直接删掉就行
del tst_user_item_feats_df['label']

返回排序后的结果

def submit(recall_df, topk=5, model_name=None):
    recall_df = recall_df.sort_values(by=['user_id', 'pred_score'])
    recall_df['rank'] = recall_df.groupby(['user_id'])['pred_score'].rank(ascending=False, method='first')
    
    # 判断是不是每个用户都有5篇文章及以上
    tmp = recall_df.groupby('user_id').apply(lambda x: x['rank'].max())
    assert tmp.min() >= topk
    
    del recall_df['pred_score']
    submit = recall_df[recall_df['rank'] <= topk].set_index(['user_id', 'rank']).unstack(-1).reset_index()
    
    submit.columns = [int(col) if isinstance(col, int) else col for col in submit.columns.droplevel(0)]
    # 按照提交格式定义列名
    submit = submit.rename(columns={
   '': 'user_id', 1: 'article_1', 2: 'article_2', 
                                                  3: 'article_3', 4: 'article_4', 5: 'article_5'})
    
    save_name = save_path + model_name + '_' + datetime.today().strftime('%m-%d') + '.csv'
    submit.to_csv(save_name, index=False, header=True)

# 排序结果归一化
def norm_sim(sim_df, weight=0.0):
    # print(sim_df.head())
    min_sim = sim_df.min()
    max_sim = sim_df.max()
    if max_sim == min_sim:
        sim_df = sim_df.apply(lambda sim: 1.0)
    else:
        sim_df = sim_df.apply(lambda sim: 1.0 * (sim - min_sim) / (max_sim - min_sim))

    sim_df = sim_df.apply(lambda sim: sim + weight)  # plus one
    return sim_df

LGB模型是GBDT模型中应用最广泛的模型，LGB节省内存且计算速度更快。

LGB排序模型

# 防止中间出错之后重新读取数据
trn_user_item_feats_df_rank_model = trn_user_item_feats_df.copy()

if offline:
    val_user_item_feats_df_rank_model = val_user_item_feats_df.copy()
    
tst_user_item_feats_df_rank_model = tst_user_item_feats_df.copy()

# 定义特征列
lgb_cols = ['sim0', 'time_diff0', 'word_diff0','sim_max', 'sim_min', 'sim_sum', 
            'sim_mean', 'score','click_size', 'time_diff_mean', 'active_level',
            'click_environment','click_deviceGroup', 'click_os', 'click_country', 
            'click_region','click_referrer_type', 'user_time_hob1', 'user_time_hob2',
            'words_hbo', 'category_id', 'created_at_ts','words_count']

# 排序模型分组
trn_user_item_feats_df_rank_model.sort_values(by=['user_id'], inplace=True)
g_train = trn_user_item_feats_df_rank_model.groupby(['user_id'], as_index=False).count()["label"].values

if offline:
    val_user_item_feats_df_rank_model.sort_values(by=['user_id'], inplace=True)
    g_val = val_user_item_feats_df_rank_model.groupby(['user_id'], as_index=False).count()["label"].values

# 排序模型定义
lgb_ranker = lgb.LGBMRanker(boosting_type='gbdt', num_leaves=31, reg_alpha=0.0, reg_lambda=1,
                            max_depth=-1, n_estimators=100, subsample=0.7, colsample_bytree=0.7, subsample_freq=1,
                            learning_rate=0.01, min_child_weight=50, random_state=2018, n_jobs= 16)  

# 排序模型训练
if offline:
    lgb_ranker.fit(trn_user_item_feats_df_rank_model[lgb_cols], trn_user_item_feats_df_rank_model['label'], group=g_train,
                eval_set=[(val_user_item_feats_df_rank_model[lgb_cols], val_user_item_feats_df_rank_model['label'])], 
                eval_group= [g_val], eval_at=[1, 2, 3, 4, 5], eval_metric=['ndcg', ], early_stopping_rounds=50, )
else:
    lgb_ranker.fit(trn_user_item_feats_df[lgb_cols], trn_user_item_feats_df['label'], group=g_train)

# 模型预测
tst_user_item_feats_df['pred_score'] = lgb_ranker.predict(tst_user_item_feats_df[lgb_cols], num_iteration=lgb_ranker.best_iteration_)

# 将这里的排序结果保存一份，用户后面的模型融合
tst_user_item_feats_df[['user_id', 'click_article_id', 'pred_score']].to_csv(save_path + 'lgb_ranker_score.csv', index=False)

# 预测结果重新排序, 及生成提交结果
rank_results = tst_user_item_feats_df[['user_id', 'click_article_id', 'pred_score']]
rank_results['click_article_id'] = rank_results['click_article_id'].astype(int)
submit(rank_results, topk=5, model_name='lgb_ranker')

# 五折交叉验证，这里的五折交叉是以用户为目标进行五折划分
#  这一部分与前面的单独训练和验证是分开的
def get_kfold_users(trn_df, n=5):
    user_ids = trn_df['user_id'].unique()
    user_set = [user_ids[i::n] for i in range(n)]
    return user_set

k_fold = 5
trn_df = trn_user_item_feats_df_rank_model
user_set = get_kfold_users(trn_df, n=k_fold)

score_list = []
score_df = trn_df[['user_id', 'click_article_id','label'</