really about to sleep. a few things.
- col 12 are declining. this was seen after creating a time graph feature in the visualizer. really hoping they rebound, even marginally. i really hope the edge effects don't take them out uniformally. that would be bad. that would be 8 more brave EB soldiers sacrificed out of the current 93, which would make them 85. please rebound bro.
- hull-derived features might want to be discarded manually during preprocessing of modeling.
- meeting w/ sun and b are mandatory (the week of the 9th) to discuss potential solutions to counter losing track of individuality during staining process. a slideshow and progress report should be brought forth. i have noted both in my calendar. and it's not hard to check back at the convo b/w me and b.
let's see what happens by day 21.