Research Backlogs - Githubissues

rahlk / Research

0 stars 0 forks source link

Research Backlogs #9

Open rahlk opened 7 years ago

rahlk commented 7 years ago

Bellwether Extension

[ ] A repair/extension to the transfer learning paper
[ ] A critique of measurement methods in CC learning (e.g. everyone measures usign X but when we unpack X in Y,Z we see ugh)

XTREES

[ ] and need more work on xtrees
[ ] that IST paper (looks good, could still fuckup but, it looks good) something BETTER than xtrees. causality? scalability?

More on causality stuff (sounds ok but ...)

[ ] how does it connect to xtrees?
[ ] if you do something simple like use it to build rules then assess those rules by
- comparing with other rule learners
- Score those causal rules via medMRE or pd,pdf

rahlk commented 7 years ago

Causality stuff (sounds ok but ...)

[X] How does it connect to xtrees? See #10.
[ ] Better XTREEs? Combine causal rules this using GENESIM.

Planned

[ ] Score those causal rules via medMRE or pd,pd

To Do

[ ] Comparing with other rule learners

rahlk commented 7 years ago

Bellwether Extension

Results - TCA+ vs. WHICH.pdf

log4j lucene