List your company
ResearchFormerly WizWand

Binary Classification on IMDB 12 LFs (test)

86.72F1 ScoreGround truth
68.041672.890877.7482.5892Jul 5, 2021
Updated 8mo ago

Evaluation results

MethodLinks
Ground truthTraining Label Source=True training labels, Number of LFs=122021.0786.72
Ground truthtraining_data=true training labels2021.0786.72
WeaSELTraining Label Source=End-to-end joint learning, Number of LFs=122021.0777.22
WeaSEL2021.0777.22
SnorkelTraining Label Source=Probabilistic labels estimated by latent label model, Number of LFs=122021.0774.45
Snorkel2021.0774.45
WeaSEL-randomencoder=randomly initialized2021.0774.22
Majority voteTraining Label Source=Hard labels induced by majority vote of the LFs, Number of LFs=122021.0774.13
Majority vote2021.0774.13
Triplet MeanTraining Label Source=Probabilistic labels estimated by latent label model, Number of LFs=122021.0773.18
Triplet-Mean2021.0773.18
TripletTraining Label Source=Probabilistic labels estimated by latent label model, Number of LFs=122021.0773.15
Triplet2021.0773.15
Triplet-medianTraining Label Source=Probabilistic labels estimated by latent label model, Number of LFs=122021.0773
Sup. (Val. set)Training Label Source=Hard labels induced by the labeled validation set, Number of LFs=122021.0768.76
Sup. (Val. set)training_data=labeled validation set2021.0768.76