Data for Machine Learning

25 Apr 2020

how do you collect data…?

how do you detect data drift, how do you regularly assess performance of the algorithm?

to prove that it works is one thing, to prove that it will not fail requires far greater effort.


after you deploy your algorithm, how do you stay close to it

how do you push updates to it?

how can you update the software?

plug in a memory card and reboot…?

why can’t/shouldn’t it be connected…?


in other news, some free ebooks for download:


i wonder how long i will take to just get through (much less digest) all of the knowledge in these books…!

just as well that i’ve lost interest in fiction (for now) LOL