### tencent_2017_contest_final http://algo.tpai.qq.com/home/information/index.html (2017腾讯社交广告算法大赛)
---------------------------------------------------------------------------------------------
## dependency
> sudo pip install lightgbm pandas numpy sklearn fire
---------------------------------------------------------------------------------------------
## useage
> ./run_phase_1.sh
##### need: 8G RAM; 50G disk; 2 hour time;
##### 初赛的脚本只需 8G 运行内存， 50G 硬盘空间， 整个流程只需不到 2 个小时
or
> ./run.sh
##### need: 8G RAM; 500G disk; 10 hour time;
##### 决赛的脚本同样只需 8G 运行内存, 500G 硬盘空间， 整个流程需要大约 10 个小时
---------------------------------------------------------------------------------------------
## notice
##### the run.sh or run_phase_1.sh will auto download the data, auto preprocess, auto extract feature, auto run model
##### run.sh 或 run_phase_1.sh 会自动下载数据，自动预处理，自动提取特征，自动跑模型生成submiss.csv.
---------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------
## other solutions
##### final 14th https://github.com/z564808896/Tencent_Social_Ads
##### final 23th https://github.com/BladeCoda/Tencent2017_Final_Coda_Allegro
第一届腾讯社交广告高校算法大赛参赛源码+项目说明.zip
版权申诉
159 浏览量
2024-01-25
18:14:41
上传
评论
收藏 472KB ZIP 举报
第一届腾讯社交广告高校算法大赛参赛源码+项目说明.zip （43个子文件） 
code_20105 
utils.py 5KB sql extract_feature.sql 34KB sql2csv.sql 486B send_email.py 1016B
feats_csvs.txt 2KB data_describe.txt 1KB old_features.csv 8KB add_one_column.py 873B extract_feature.py 13KB datasets train Readme.md 0B tb_2.csv 32B sort.sh 68B tb_12.csv 410B tb_1.csv 398B temp.sql 763B submit Readme.md 0B join.sh 100B read_csv_to_sql.py 2KB dataset Readme.md 0B use_join.py 1KB read_all_csv_to_sql.py 154B join_one_key.sh 390B tecent_use_sort.sh 402B use_sort.py 671B tecent_use_join.sh 3KB read_csv_to_sql_special.py 3KB join_two_key.sh 516B sort_then_join.sh 363B feats Readme.md 50B features_after_corr_select.txt 5KB features.csv 16KB merge_use_small_dtype.py 2KB split_data.py 3KB extract_other_feature.py 4KB run_phase_1.sh 2KB run.sh 2KB fillna_for_final_csv.py 884B TSA_history.py 5KB feature_name_importance.txt 3.37MB model_single.py 17KB feat_select_by_corr.py 2KB Readme.md 1KB tf_LR.py 5KB资源评论
