<aside>
〰️ terminology
- [x] 추상적인(전문적인) 용어임.
- [x] 키워드가 가까운 시일 내 검색될 것 같지 않음.
- [x] 알아두고 싶은 단어이지만 영구노트 생성은 과분함.
</aside>
정의
39:50, Data flywheel is the idea that for certain machine learning projects, you can create this positive feedback cycle where as you improve your model that improvement in your model performance leads to a better product leads to more users. The more users you have the more data you’re able to collect …
[[[SW Environment](https://www.notion.so/Environment-d73d08b95a824b3f957eb0134f8ced1f) MLOps](https://www.notion.so/MLOps-1be9e881d5444f268fde8c1ea7e00f59) Feedback Loop](https://www.notion.so/Feedback-Loop-4ea8b9c67af9461f9e51a18a28032993) 과 완전히 동일한 단어임.
이 단어를 발견한 곳
- Aggregate, process, clean, label, and version data, write and debug model code, provision compute (somehow find compute), run experiments, review results, deploy model, monitor predictions and close data flywheel loop.