Hacker News Clone new | comments | show | ask | jobs | submit | github repologin
Google Rules of Machine Learning (2018) (developers.google.com)
33 points by pongogogo 2 hours ago | hide | past | web | 5 comments | favorite





> if you are ranking apps in an app marketplace, you could use the install rate or number of installs as heuristics. If you are detecting spam, filter out publishers that have sent spam before. Don’t be afraid to use human editing either. If you need to rank contacts, rank the most recently used highest (or even rank alphabetically). If machine learning is not absolutely required for your product, don't use it until you have data.

Every one of these suggested heuristics is derived from data, except for alphabetical order. It’s interesting to me that the author uses the word ‘data’ to mean ‘lots of examples’: a rather narrow meaning.


This appears to be pretty old (it has a reference to Google Plus) but seems like good generic advice.

Should say "(2018)".

A rare example of practices from Google which would be transferable to other organizations.

Rule 1 is golden and oft forgotten.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: