search

ntucllab / libact

Pool-based active learning in Python
http://libact.readthedocs.org/
BSD 2-Clause "Simplified" License
779 stars 174 forks source link

density_weighted_uncertainty_sampling #72

Closed yangarbiter closed 7 years ago

yangarbiter commented 7 years ago

This active learner requires predict_proba not predict_real for libact.models.

I personally favor interface like scikit-learn, instead of dividing them to Model, ContinuousModel, ProbabilisticModel.

Discussion for interface needed.

@hsuantien @iamyuanchung

codecov-io commented 7 years ago

Current coverage is 86.53% (diff: 98.71%)

Merging #72 into master will increase coverage by 1.58%

@@             master        #72   diff @@
==========================================
  Files            20         21     +1   
  Lines           598        676    +78   
  Methods           0          0          
  Messages          0          0          
  Branches          0          0          
==========================================
+ Hits            508        585    +77   
- Misses           90         91     +1   
  Partials          0          0

Powered by Codecov. Last update c40302e...f265f7f

iamyuanchung commented 7 years ago

I personally think that we should better follow the design philosophy of sklearn, as most of the users are expected to be already familiar with it.