一个完整的机器学习项目生产环境部署.zip

## 项目说明 - 一个完整的二分类机器学习项目生产环境部署,包括接收请求原始变量入参,<br> 参数检测, 数据预处理, 模型变量构造, 模型预测, 数据库操作, 返回结果. <br> - API入参 ```python { "order_id":"123456789", "data":{ "old_customer":"0", "debt_amount":"7000000", "debt_number":"6", "recent24hours_apply":"1", "recent7days_apply":"13", "recent30days_apply":"31", "debt_level":"GENERAL", "education":"S1", "iswhatsapp":"yes", "whatsapp_avatar":"yes", "whatsapp_company_account":"no", "idinquiriesuname_21d":"9", "idinquiriesuname_30d":"11", "topup_360_720_max":"100", "topup_360_720_avg":"32", "topup_0_360_min":"1", "topup_180_360_avg":"51", "preference_bank_60d":"0", "A_II_360d":"20", "A_PI_360d":"11" } } ``` - API出参 ```python { "created_time": "2020-02-24 23:27:49", "ml_score": 768, "order_id": "测试订单", "reason": "请求成功", "status": 200, "version": "0.0.1" } ``` *** ### 文件 - __web__ &nbsp;&nbsp;web层 - __util__ &nbsp;&nbsp;通用工具 - __resources__ &nbsp;&nbsp;资源配置 - __general_model__ &nbsp;&nbsp;逻辑层 - __general_model.model__ &nbsp;&nbsp;数据预处理和模型预测 - __general_model.dao__ &nbsp;&nbsp;数据库操作 - __general_model.data__ &nbsp;&nbsp;模型文件 - __start.sh__ &nbsp;&nbsp;启动服务 - __stop.sh__ &nbsp;&nbsp;关闭服务 ### 服务部署 #### step1: 环境配置 1. 系统环境: centOS7.0+或ubuntu16.0+, 4G+内存 2. 安装anaconda3 <br> 安装包: Anaconda3-2019.03-Linux-x86_64.sh <br> 安装: bash Anaconda3-2019.03-Linux-x86_64.sh 3. 安装依赖 <br> 依赖文件: requirements.txt <br> 安装: pip install -r requirements.txt #### step2: 启动/关闭服务 1. 启动服务命令: bash start.sh 2. 关闭服务命令: bash stop.sh