IJCAI-18alimama:IJCAI-18 阿里妈妈搜索广告转化预测大赛，top50方案-源码
IJCAI-18 阿里妈妈搜索广告大赛
赛题介绍
给定广告点击相关的用户（user）、广告商品（ad）、检索词（query）、上下文内容（context）、商店（shop）等信息的条件下预测广告产生购买行为的概率（pCVR）。其中初赛预测的是普通日期的广告转化率，复赛预测的是特殊日期的广告转化率。
详细信息：
最终成绩：复赛48名
赛题分析
本赛题的正负样本比例极度不均衡，但是因为采用的是logloss评价函数，不适合抽样更改训练数据集的正负样本比，同样的有些模型的参数（比如xgb的scale_pos_weight)也是不宜用的，这些在模型的官方文档上也有说明。
单一用户的记录太少，很多用户在整个数据集中都只有一条记录，因此用户维度的特征比较难挖掘，相反商品维度的特征会更有帮助一些。
初赛与复赛由于预测的日期属性不同，造成数据分布上很大的变化。原本初赛的特征方法和模型构建方法仅仅是对复赛有
文件下载
资源详情
[{"title":"（ 74 个子文件 633.29MB ） IJCAI-18alimama:IJCAI-18 阿里妈妈搜索广告转化预测大赛，top50方案-源码","children":[{"title":"IJCAI-18alimama-master","children":[{"title":"README.md <span style='color:#111;'> 8.95KB </span>","children":null,"spread":false},{"title":"lr","children":[{"title":"lr2B_hashour.txt <span style='color:#111;'> 49.92MB </span>","children":null,"spread":false},{"title":"lr2B_oof_train.csv <span style='color:#111;'> 41.34MB </span>","children":null,"spread":false},{"title":"lr2A(14482).txt <span style='color:#111;'> 19.97MB </span>","children":null,"spread":false},{"title":"lr2B(14143).txt <span style='color:#111;'> 46.51MB </span>","children":null,"spread":false},{"title":"lr2A_hashour.txt <span style='color:#111;'> 21.37MB </span>","children":null,"spread":false},{"title":"lr2B.pkl <span style='color:#111;'> 546.70KB </span>","children":null,"spread":false},{"title":"lr2A.pkl <span style='color:#111;'> 504.62KB </span>","children":null,"spread":false},{"title":"lr2_adjust_result.py <span style='color:#111;'> 2.60KB </span>","children":null,"spread":false},{"title":"lr2A_oof_train.csv <span style='color:#111;'> 41.33MB </span>","children":null,"spread":false},{"title":"lr2A_oof_test.csv <span style='color:#111;'> 19.89MB </span>","children":null,"spread":false},{"title":"lr2.py <span style='color:#111;'> 33.40KB </span>","children":null,"spread":false},{"title":"lr2B_oof_test.csv <span style='color:#111;'> 46.47MB </span>","children":null,"spread":false}],"spread":false},{"title":"ensemble","children":[{"title":"blending2B(14002).txt <span style='color:#111;'> 46.56MB </span>","children":null,"spread":false},{"title":"blending2B(13961).txt <span style='color:#111;'> 46.59MB </span>","children":null,"spread":false},{"title":"blending2.py <span style='color:#111;'> 1.97KB </span>","children":null,"spread":false}],"spread":true},{"title":"fea_factory2.py <span style='color:#111;'> 49.06KB </span>","children":null,"spread":false},{"title":"keng","children":[{"title":"fusai_b_xgb_5_12_adjust.txt <span style='color:#111;'> 46.58MB </span>","children":null,"spread":false},{"title":"fusai_b_xgb_5_14_normal_adjust.txt <span style='color:#111;'> 46.59MB </span>","children":null,"spread":false},{"title":"fusai_b_xgb_5_15_adjust.txt <span style='color:#111;'> 46.58MB </span>","children":null,"spread":false}],"spread":true},{"title":"fm","children":[{"title":"fm2.py <span style='color:#111;'> 38.08KB </span>","children":null,"spread":false}],"spread":true},{"title":"ffm","children":[{"title":"ffm2.py <span style='color:#111;'> 44.28KB </span>","children":null,"spread":false},{"title":"ffm1A(8271).txt <span style='color:#111;'> 591.60KB </span>","children":null,"spread":false},{"title":"ffm1.py <span style='color:#111;'> 33.00KB </span>","children":null,"spread":false},{"title":"libffm","children":[{"title":"libffm_from_github <span style='color:#111;'> 0B </span>","children":null,"spread":false}],"spread":true},{"title":"ffm1A(8192).txt <span style='color:#111;'> 693.00KB </span>","children":null,"spread":false},{"title":"ffm1A_oof_train.csv <span style='color:#111;'> 15.06MB </span>","children":null,"spread":false},{"title":"ffm1A_oof_test.csv <span style='color:#111;'> 628.04KB </span>","children":null,"spread":false}],"spread":true},{"title":"lake","children":[{"title":"lgb2B_with_normal_hashour_after_lake_5_15_9_all.txt <span style='color:#111;'> 41.45MB </span>","children":null,"spread":false}],"spread":true},{"title":"data","children":[{"title":"put_dataset_here <span style='color:#111;'> 0B </span>","children":null,"spread":false}],"spread":true},{"title":"xgb1","children":[{"title":"xgboost1B.model <span style='color:#111;'> 552.12KB </span>","children":null,"spread":false},{"title":"xgboost1A_oof_test(8168).txt <span style='color:#111;'> 728.86KB </span>","children":null,"spread":false},{"title":"xgboost2B_oof_test.csv <span style='color:#111;'> 46.47MB </span>","children":null,"spread":false},{"title":"xgboost1A(8119).txt <span style='color:#111;'> 729.46KB </span>","children":null,"spread":false},{"title":"xgboost2B_hashour.txt <span style='color:#111;'> 49.93MB </span>","children":null,"spread":false},{"title":"xgboost1A(8157).txt <span style='color:#111;'> 729.09KB </span>","children":null,"spread":false},{"title":"xgboost2B(14054).txt <span style='color:#111;'> 46.53MB </span>","children":null,"spread":false},{"title":"xgboost2B.txt <span style='color:#111;'> 46.60MB </span>","children":null,"spread":false},{"title":"xgboost2B(merge13975).txt <span style='color:#111;'> 46.59MB </span>","children":null,"spread":false},{"title":"xgboost1B(8495).txt <span style='color:#111;'> 1.66MB </span>","children":null,"spread":false},{"title":"xgboost2A(14711).txt <span style='color:#111;'> 19.92MB </span>","children":null,"spread":false},{"title":"xgboost2A(14517).txt <span style='color:#111;'> 20.00MB </span>","children":null,"spread":false},{"title":"xgboost1A(8137).txt <span style='color:#111;'> 728.87KB </span>","children":null,"spread":false},{"title":"xgboost2B_add_normal_oof_test.csv <span style='color:#111;'> 46.48MB </span>","children":null,"spread":false},{"title":"xgboost1A.model <span style='color:#111;'> 366.29KB </span>","children":null,"spread":false},{"title":"xgboost2B_oof_train.csv <span style='color:#111;'> 41.37MB </span>","children":null,"spread":false},{"title":"xgboost2B(14054)_adjust_mean.txt <span style='color:#111;'> 46.58MB </span>","children":null,"spread":false},{"title":"xgboost1A-1(8157).txt <span style='color:#111;'> 728.98KB </span>","children":null,"spread":false},{"title":"xgboost1A(8149).txt <span style='color:#111;'> 729.15KB </span>","children":null,"spread":false},{"title":"xgboost1B_oof_test.csv <span style='color:#111;'> 1.66MB </span>","children":null,"spread":false},{"title":"xgboost2B_15normal.txt <span style='color:#111;'> 46.61MB </span>","children":null,"spread":false},{"title":"xgboost2B_add_normal(merge13975).txt <span style='color:#111;'> 41.44MB </span>","children":null,"spread":false},{"title":"xgboost2B.model <span style='color:#111;'> 642.15KB </span>","children":null,"spread":false},{"title":"xgboost1A(8107).txt <span style='color:#111;'> 728.76KB </span>","children":null,"spread":false},{"title":"xgboost1B_oof_train.csv <span style='color:#111;'> 15.51MB </span>","children":null,"spread":false},{"title":"xgboost2A.model <span style='color:#111;'> 739.75KB </span>","children":null,"spread":false},{"title":"xgboost1A(8168).txt <span style='color:#111;'> 728.80KB </span>","children":null,"spread":false},{"title":"xgboost2_adjust_result.py <span style='color:#111;'> 2.81KB </span>","children":null,"spread":false},{"title":"xgboost2A_post_pro(14514).txt <span style='color:#111;'> 17.79MB </span>","children":null,"spread":false},{"title":"xgboost2B_add_normal.model <span style='color:#111;'> 879.70KB </span>","children":null,"spread":false},{"title":"xgboost1B(8488).txt <span style='color:#111;'> 1.66MB </span>","children":null,"spread":false},{"title":"xgboost1A(8108).txt <span style='color:#111;'> 729.08KB </span>","children":null,"spread":false},{"title":"xgboost1.py <span style='color:#111;'> 53.39KB </span>","children":null,"spread":false},{"title":"xgboost2A_oof_test.csv <span style='color:#111;'> 19.92MB </span>","children":null,"spread":false},{"title":"xgboost2A(14449).txt <span style='color:#111;'> 19.98MB </span>","children":null,"spread":false},{"title":"xgboost2A_hashour.txt <span style='color:#111;'> 21.41MB </span>","children":null,"spread":false},{"title":"xgboost2B_add_normal_oof_train.csv <span style='color:#111;'> 64.02MB </span>","children":null,"spread":false},{"title":"xgboost2A_oof_train.csv <span style='color:#111;'> 41.33MB </span>","children":null,"spread":false},{"title":"xgboost2A(14376).txt <span style='color:#111;'> 19.98MB </span>","children":null,"spread":false},{"title":"xgboost2B_add_normal_hashour.txt <span style='color:#111;'> 49.94MB </span>","children":null,"spread":false},{"title":"xgboost2.py <span style='color:#111;'> 20.80KB </span>","children":null,"spread":false},{"title":"xgboost1A_oof_test.csv <span style='color:#111;'> 728.65KB </span>","children":null,"spread":false},{"title":"xgboost1A_oof_train.csv <span style='color:#111;'> 15.48MB </span>","children":null,"spread":false},{"title":"xgboost2A(1442).txt <span style='color:#111;'> 19.98MB </span>","children":null,"spread":false}],"spread":false}],"spread":true}],"spread":true}]
评论信息
其他资源
- 相当标准的VRML校园漫游和公路汽车（动画）程序
- 对chunked编码进行解码
- 高阶累积量 matlab
- ELM极限学习机 Matlab代码
- div+css实现网上书店网页
- stm32的DXP元件库
- turbo码译码过程
- PCI9056经典调试记录
- 复杂网络基础6.ppt
- 最全省市区三级联动excel
- YS-H7Multi_HAL-143. 内部SRAM使用.rar
- HC-SR04超声波测距模块+OLED12864显示
- ITIL全套中文版
- K3 10.3破解文件
- 技术白皮书模板
- 中南大学最优控制资料（英文课件）
- Spotlistr：:headphone:将任何内容转换为Spotify播放列表！-源码
- 51单片机自学笔记（6寸pdf）
- Dreamwever个人主页设计
- C#五子棋
- WebService加密 for SoapHeader
- 基于视觉的AGV转向模型和控制研究
- ARUBA控制器上的常用命令
免责申明
【只为小站】的资源来自网友分享，仅供学习研究，请务必在下载后24小时内给予删除，不得用于其他任何用途，否则后果自负。基于互联网的特殊性，【只为小站】 无法对用户传输的作品、信息、内容的权属或合法性、合规性、真实性、科学性、完整权、有效性等进行实质审查；无论 【只为小站】 经营者是否已进行审查，用户均应自行承担因其传输的作品、信息、内容而可能或已经产生的侵权或权属纠纷等法律责任。
本站所有资源不代表本站的观点或立场，基于网友分享，根据中国法律《信息网络传播权保护条例》第二十二条之规定，若资源存在侵权或相关问题请联系本站客服人员，zhiweidada#qq.com，请把#换成@，本站将给予最大的支持与配合，做到及时反馈和处理。关于更多版权及免责申明参见 版权及免责申明
本站所有资源不代表本站的观点或立场，基于网友分享，根据中国法律《信息网络传播权保护条例》第二十二条之规定，若资源存在侵权或相关问题请联系本站客服人员，zhiweidada#qq.com，请把#换成@，本站将给予最大的支持与配合，做到及时反馈和处理。关于更多版权及免责申明参见 版权及免责申明