Binary LR samples using T4 templates #3099
Conversation
Codecov Report
@@ Coverage Diff @@
## master #3099 +/- ##
==========================================
+ Coverage 72.52% 72.52% +<.01%
==========================================
Files 808 808
Lines 144665 144665
Branches 16198 16198
==========================================
+ Hits 104913 104916 +3
+ Misses 35342 35338 -4
- Partials 4410 4411 +1
|
Codecov Report
@@ Coverage Diff @@
## master #3099 +/- ##
==========================================
- Coverage 72.52% 72.51% -0.01%
==========================================
Files 808 808
Lines 144665 144665
Branches 16198 16198
==========================================
- Hits 104913 104907 -6
- Misses 35342 35346 +4
- Partials 4410 4412 +2
|
| Label = label, | ||
| // Create random features that are correlated with the label. | ||
| // For data points with false label, the feature values are slightly increased by adding a constant. | ||
| Features = Enumerable.Repeat(label, 50).Select(x => x ? randomFloat() : randomFloat() + 0.03f).ToArray() |
There was a problem hiding this comment.
0.03f [](start = 108, length = 5)
The accuracy is low. We want samples to ideally have 80%+ accuracy. Can we increase this to 0.1 and see how much accuracy improves. If so, let's keep tree samples with old 0.03f value, and increase it for less powerful trainers like LogisticRegression. #Resolved
|
|
||
| namespace Microsoft.ML.Samples.Dynamic.Trainers.BinaryClassification | ||
| { | ||
| public static class LogisticRegressionWithOptions |
There was a problem hiding this comment.
LogisticRegressionWithOptions [](start = 24, length = 29)
please make sure the sample is added to the xml documentation for BinaryClassification.Trainers.LbfgsLogisticRegression. #Resolved
| @@ -0,0 +1,29 @@ | |||
| <#@ include file="BinaryClassification.ttinclude"#> | |||
| <#+ | |||
| string ClassName="LogisticRegression"; | |||
There was a problem hiding this comment.
LogisticRegression [](start = 18, length = 18)
let's rename class to LbfgsLogisticRegression to match the API. also plz rename the other sample to LbfgsLogisticRegressionWithOptions #Resolved
|
Thanks! |
ghost
locked as resolved and limited conversation to collaborators
Related to #2522. The *.cs files are auto-generated. Please review the .tt and .ttinclude files.