stevenae

The quantitative ux research team at Google was created for exactly this problem: a service which became popular before the right metrics existed, meaning metrics need to be derived first, then optimized. We would observe users (irl), read their logs, then generate experiments to improve the behavior as measured by logs, and return to see if the experiment improves irl experiences. There were not many of us and we are around :)

My reading of this situation is that MAPE would do the opposite. Means are skewed towards outliers.

Thanks for the reply! I am outside the forecasting sphere.

RMSLE gives proportional error (so, scale-invariant) without MAPE's systematic under-prediction bias. It does require all-positive values, for the logarithm step.

To clarify, you'd prefer rmsle?

For this and sibling -- yes. Essentially, using the output of any model as an input to another model is transfer learning.

> Lately, I just steal embeddings from big models and slap a dumb classifier on top. Works better, runs faster, less drama.

You may know this but many don't -- this is broadly known as "transfer learning".

Thank you!

How accurate is his claim that Augustus became emperor through (my paraphrasing) democratic means and promises to fix real problems for Romans?

This still strikes me as escapism.

There was a saying at Google, I code for free, they pay me for XYZ (literally everything else).

I guess my quibble is with the percentage, then. A good, cheap, plentiful camera belies the idea that only the top 0.1% of cameras were good.

Disagree with the first piece about only using the top 0.1%. I grew up (through my 20's) shooting on a Pentax K1000, cheap workhorse of a camera, and I preferred its ergonomics to top-end mirrorless cameras I use today.