{"id":19750193,"url":"https://github.com/wepe/ppd_riskcontrolcompetition","last_synced_at":"2025-04-06T10:13:40.859Z","repository":{"id":87470016,"uuid":"59303327","full_name":"wepe/PPD_RiskControlCompetition","owner":"wepe","description":"3rd Place Solution for PPD Risk Control Competition","archived":false,"fork":false,"pushed_at":"2017-03-20T14:59:47.000Z","size":909,"stargazers_count":340,"open_issues_count":5,"forks_count":249,"subscribers_count":35,"default_branch":"master","last_synced_at":"2025-03-30T09:07:44.436Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Java","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/wepe.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2016-05-20T14:58:22.000Z","updated_at":"2024-12-13T02:41:37.000Z","dependencies_parsed_at":null,"dependency_job_id":"a6ff8729-d982-4ef5-b24e-2b49e3bcfb0a","html_url":"https://github.com/wepe/PPD_RiskControlCompetition","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wepe%2FPPD_RiskControlCompetition","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wepe%2FPPD_RiskControlCompetition/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wepe%2FPPD_RiskControlCompetition/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/wepe%2FPPD_RiskControlCompetition/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/wepe","download_url":"https://codeload.github.com/wepe/PPD_RiskControlCompetition/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247464224,"owners_count":20942970,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-11-12T02:31:20.786Z","updated_at":"2025-04-06T10:13:40.829Z","avatar_url":"https://github.com/wepe.png","language":"Java","funding_links":[],"categories":[],"sub_categories":[],"readme":"### [拍拍贷风险控制大赛铜奖解决方案](https://github.com/wepe/PPD_RiskControlCompetition)\n\n- 赛题介绍\n\n\t请参见 [Kesci官网介绍](http://www.kesci.com/apps/home_log/index.html#!/competition/56cd5f02b89b5bd026cb39c9)\n\n- 参赛队伍\n\n\t不得仰视本王\n\n\t队员: [天音](https://tianchi.shuju.aliyun.com/science/scientistDetail.htm?spm=5176.100065.0.0.gnyTyY\u0026userId=713068774)，[Cyber](https://tianchi.shuju.aliyun.com/science/scientistDetail.htm?spm=0.0.0.0.hITk7I\u0026userId=728128096)，[Bryan](http://blog.csdn.net/bryan__)，[赵蕊](https://tianchi.shuju.aliyun.com/science/scientistDetail.htm?spm=5176.100170.111.28.PytD5w\u0026userId=312633256)，[wepon](http://2hwp.com/)\n\n- 解决方案\n\n\t详细解决方案请看 [PDF文件](https://github.com/wepe/PPD_RiskControlCompetition/blob/master/%E9%A3%8E%E6%8E%A7%E7%AE%97%E6%B3%95%E5%A4%A7%E8%B5%9B%E8%A7%A3%E5%86%B3%E6%96%B9%E6%A1%88.pdf)\n\n- 代码目录说明\n\n\t- `proccess  文件夹`\n\t\n\t\t- Split.java将采样数据分块\n\t\t- CombineSample.java将采样数据合并\n\t\t- NThreadRNB.java和NThreadMatrix.java是并行采样代码工具类\n\t\n\t\n\t- `feature engineering  文件夾`\n\t\n\t\t- NullDiscrete.java和CityFeature.java为计算城市向量特征\n\t\t- city_feature.sql 为数据库处理代码\n\t\t- rank.py 对原始数值特征进行排序，得到排序特征\n\t\t- null.py 分析和处理缺失值\n\t\t- jw.py  生成经纬度特征\n\t\t- CategoryFeatureProcess.py\t特征处理\n\t\t- MergeFeature.py\t特征合并\n\t\t- SelectFeature.py\t特征筛选\n\t\t- jingch文件夹\n\t\t\t- UserLogInfoFeature\t 登录信息特征提取类\n\t\t\t- UserUpdateInfoFeature\t修改信息特征提取类\n\t\t\t- MergeTool\t模型融合工具类\n\t\t\t- FeatureProcess\t特征处理类\n\t\n\t\n\t\n\t- `feature_select 文件夹`\n\t\n\t\t- sort_feature_using_xgb.py  训练xgb模型对特征进行重要性排序，特征选择\n\t\t- avg_featurescore.py 将多份featurescore文件全加，得到特征重要性排序文件\n\t\n\t\n\t\n\t\n\t- `sample  文件夹`\n\t\n\t\t包含了两种采样代码，其中MY0为本次比赛使用的基于粗糙集的并行过采样算法\n\t\n\t\n\t- `util  文件夹`\n\t\n\t\t  这个文件夹包含一些常用的工具类和代码\n\t\t- visualize_null.py 缺失值的可视化分析\n\t\t- visualize_dataset.py 对每日成交量和违约量进行可视化\n\t\t- plot_feature_importance.py 画特征重要性图\n\t\t- cal_auc.py 计算auc，线下验证\n\t\n\t\n\t- `lr文件夹`\n\t\n\t\t- lr.py  逻辑回归模型文件\n\t\n\t\n\t\n\t- `svm 文件夹`\n\t\n\t\t- svm.py  运用数据集分解的方法训练多个svm进行averaging\n\t\t- avg_val.py  查看在验证集上的效果\n\t\t- avg_test.py 查看在测试集上的效果\n\t\n\t\n\t- `xgb文件夹`\n\t\n\t\t- single_xgb.py  单模型xgboost，线下cv的auc值是0.782左右\n\t\t- bagging_xgb.py  单模型xgboost的改进版，对参数和特征加入随机扰动，训练多个xgboost子模型\n\t\t- avg_test.py  对子模型预测结果进行averaging\n\t    - graphlab_xgboost.py graphlab版本的xgboost\n\t\n\t- `ensemble文件夹`\n\t\n\t\t- cal_mic.py 计算单模型结果文件之间的相关性\n\t\t- ensemble.py 多个单模型的加权融合,通过验证集选取最优参数\n\t\t- blend_ensemble.py 训练多个单模型，并进行blending融合\n\t\n\n- 使用须知\n\n\t- 代码可以自由使用，但请保留出处。\n\n\n\t\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fwepe%2Fppd_riskcontrolcompetition","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fwepe%2Fppd_riskcontrolcompetition","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fwepe%2Fppd_riskcontrolcompetition/lists"}