ML Day #5 — Udacity Course Cont.

• Learned some stuff about ndimage — scipy is so cooL!
• Finished assignment #1 except for the last few problems. Will save them for later — they seem really hard. Somehow my label array ended up empty? Not sure why.
• Listened to a bit more Udacity course — highlight is that if you have a training set, take 10% of it and make it test data for validation, then 10% of that and make it “final test data” for when you think you’ve fully validated. This seems like an attempt to “get above yourself,” and there are probably deeper questions to answer here, but whatever.
• Started reading through more scipy documentation
• Continued in coursera course — they’re going over linear regression and multi-dimensional linear regression models right now. Boring stuff, but useful background/review.