Kagami Hiiragi
|
0956e40d34
|
Fix stupid bug with comparing threshold
It was working backwards.
Fixes #33 #16
|
2019-08-18 12:30:04 +03:00 |
|
Kagami Hiiragi
|
ef56f89511
|
Simplify classify
We don't need additional level of indirection and can store category in
distances vector right away.
|
2019-08-18 11:43:46 +03:00 |
|
Kagami Hiiragi
|
84d557bef3
|
Fix classify when threshold was specified
Fixes #31
|
2019-08-18 11:06:03 +03:00 |
|
Kagami Hiiragi
|
4d258ca248
|
Add ClassifyThreshold
Fixes #9
|
2019-01-14 00:11:13 +03:00 |
|
Kagami Hiiragi
|
6b3e430c90
|
Use same descriptor type everywhere
It's not possible to use fixed size with loss_metric so use arbitrary
sized matrix. Explicit size should be more effecient but not much.
|
2018-08-16 00:03:07 +03:00 |
|
Kagami Hiiragi
|
147cb0c648
|
Pass cats map as const reference
|
2018-08-13 21:18:40 +03:00 |
|
Kagami Hiiragi
|
d849158197
|
Fix style
|
2018-04-25 12:12:20 +03:00 |
|
Kagami Hiiragi
|
acafdf1721
|
Actually respect distance ordering
|
2018-04-15 03:18:27 +03:00 |
|
Kagami Hiiragi
|
b550cf37e6
|
Improve classification even more
Use ordered map so max_element will respect distance ordering.
|
2018-04-15 02:05:58 +03:00 |
|
Kagami Hiiragi
|
08327fa6fb
|
Improve classification
|
2018-04-14 19:16:48 +03:00 |
|
Kagami Hiiragi
|
894891e36f
|
Optimizations
|
2018-04-07 14:55:49 +03:00 |
|
Kagami Hiiragi
|
f94d7de55e
|
Add simple classifier
|
2018-04-06 17:18:06 +03:00 |
|