Giters

jmschrei / apricot

apricot implements submodular optimization for the purpose of selecting subsets of massive data sets to train machine learning models quickly. See the documentation page: https://apricot-select.readthedocs.io/en/latest/index.html

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Any plan to implement streaming submodular implementation for the saturated coverage problem?

badrisnps opened this issue · comments

commented

"Streaming optimization is implemented for mixtures of functions, but not for sum redundancy or saturated coverage functions."

I was wondering if there is a technical reason for the above or just that its doable but not yet implemented. Specifically I was interested in streaming sub-modular optimization for saturated coverage.

commented

I don't remember there being a reason that it couldn't be implemented. I think that it took longer than expected for me to add in streaming optimization and I needed to move on to the next project after getting a few of the more-common functions working. Unfortunately, I don't currently have plans to expand streaming optimization unless there is huge demand. Sorry. :(

commented

Thanks! If I find that I need it, perhaps I can review your current implementation for pointers and submit a Pull Request.

commented

If you happen to implement streaming submodularity for those functions I'd love to review a PR!