FinML - Model testing and monitoring with Laszlo and Neal
A wonderful evening to all of you and it was great to see so many of you at our last session, where we’ve talked with Kevin Stumpf about feature stores.
For those of you that have missed is - no worries, you can find the recording, presentation and notes here.
I’m really excited to announce to announce our next meetup, where we’ll have a roundtable discussion with Laszlo from Hypergolic and Neal from Monzo about testing and monitoring approaches for ML models. Some topics include:
How to test models? Which parts should be tested with unit tests or regression tests?
How to do exploratory testing with ML? How do you collect qualitative feedback to feed it back into model monitoring?
What are sensible tests for the data used in ML? Is range and summary statistics sufficient? Should we use the same tests for model monitoring as well or try to dynamically identify change points in distributions?
What is fairness / bias testing? How can it be applied within the LM context?
Mark your calendars for
7th of April
18:00 GMT
Sign up to this stubstack to be invited
As always - share the word by clicking the share button below and let me know if you’d like to participate in one of our future meetups!