sunnotes / Ali Data Mining
阿里巴巴大数据竞赛
Stars: ✭ 62
Labels
Projects that are alternatives of or similar to Ali Data Mining
Objectivesql
ObjectiveSQL is an ORM framework in Java based on ActiveRecord pattern, which encourages rapid development and clean, codes with the least and convention over configuration.
Stars: ✭ 1,109 (+1688.71%)
Mutual labels: sql
Aspnetcorenlog
ASP.NET Core NLog MS SQL Server PostgreSQL MySQL Elasticsearch
Stars: ✭ 54 (-12.9%)
Mutual labels: sql
Interference
opensource distributed database with base JPA implementation and event processing support
Stars: ✭ 57 (-8.06%)
Mutual labels: sql
Ddlparse
DDL parase and Convert to BigQuery JSON schema and DDL statements
Stars: ✭ 52 (-16.13%)
Mutual labels: sql
Rpg Boilerplate
Relay (React), Postgres, and Graphile (GraphQL): A Modern Frontend and API Boilerplate
Stars: ✭ 62 (+0%)
Mutual labels: sql
Jasync Sql
Java & Kotlin Async DataBase Driver for MySQL and PostgreSQL written in Kotlin
Stars: ✭ 1,092 (+1661.29%)
Mutual labels: sql
Reform
A better ORM for Go, based on non-empty interfaces and code generation.
Stars: ✭ 1,103 (+1679.03%)
Mutual labels: sql
Simple 1c
Транслятор запросов и Linq-провайдер для 1С-Бухгалтерии
Stars: ✭ 56 (-9.68%)
Mutual labels: sql
Fluent
Vapor ORM (queries, models, and relations) for NoSQL and SQL databases
Stars: ✭ 1,071 (+1627.42%)
Mutual labels: sql
Data Science Best Resources
Carefully curated resource links for data science in one place
Stars: ✭ 1,104 (+1680.65%)
Mutual labels: sql
Squid
🦑 Provides SQL tagged template strings and schema definition functions.
Stars: ✭ 57 (-8.06%)
Mutual labels: sql
Osql Experimental
A community-oriented fork of osquery with support for cmake, public CI testing, and regular releases
Stars: ✭ 62 (+0%)
Mutual labels: sql
Ali_Data_Mining
阿里巴巴大数据竞赛
简介
在天猫,每天都会有数千万的用户通过品牌发现自己喜欢的商品,品牌是联接消费者与商品最重要的纽带。本届赛题的任务就是根据用户4个月在天猫的行为日志,建立用户的品牌偏好,并预测他们在将来一个月内对品牌下商品的购买行为。
数据类型
- 字 段 字段说明 提取说明
- user_id 用户标记 抽样&字段加密
- Time 行为时间 精度到天级别&隐藏年份
- action_type 用户对品牌的行为类型 包括点击、购买、加入购物车、收藏4种行为 (点击:0 购买:1 收藏:2 购物车:3)
- brand_id 品牌数字ID 抽样&字段加密
用户对任意商品的行为都会映射为一行数据。其中所有商品ID都已汇总为商品对应的品牌ID。用户和品牌都分别做了一定程度的数据抽样,且数字ID都做了加密。所有行为的时间都精确到天级别(隐藏年份)。
评估指标
最后我们用F1-Score 来拟合准确率与召回率,并且大赛最终的比赛成绩排名以F1得分为准。
实现算法
-
协同过滤
简单建模,效果很差 -
逻辑回归
最好成绩:1111/0.06480648/0.05224964/5.785456 -
随机森林
最好成绩:1160/0.05344828/0.04499274/4.885737 -
贝叶斯 -ing
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].