知识卡片
数据到部署的可反馈流程
流程以问题和数据为基础,经预处理、拟合与评价才走向使用;不达目标时回到有问题的环节。图是学习用职责整理,实际反馈不能靠反复查看最终测试来调参。
结构图:
flowchart TD
Q[问题和验收标准] --> D[采集 探索及检查数据]
D --> P[训练内拟合预处理与特征]
P --> T[拟合和验证选择]
T --> E[独立评价最终方案]
E -->|满足条件| U[部署使用及监测]
T -->|发现问题| Q
U -->|环境或需求变化| Q
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- · 查看原书依据
The process begins with a problem definition and statement. Next, data collection and integration take place. This is followed by data exploration. The subsequent step is data preparation and cleaning. Data transformation and standardization follow. Problem formulation and exploratory data analysis, or EDA, including labeling and feature engineering, are performed. Data visualization is then conducted. The process moves to model selection, where an appropriate learning
参考来源
- 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第03章《机器学习流程与数据管理》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/14_chapter03.xhtml)
- 结论依据:摘录中关于数据到部署的可反馈流程的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:The process begins with a problem definition and statement. Next, data collection and integration take place. This is followed by data exploration. The subsequent step is data preparation and cleaning. Data transformation and standardization follow. Problem formulation and exploratory data analysis, or EDA, including labeling and feature engineering, are performed. Data visualization is then conducted. The process moves to model selection, where an appropriate learning