Uncertainty in Artificial Intelligence
First Name   Last Name   Password   Forgot Password   Log in!
    Proceedings         Authors   Author's Info   Article details         Search    
Estimating Accuracy from Unlabeled Data
Emmanouil Platanios, Avrim Blum, Tom Mitchell
We consider the question of how unlabeled data can be used to estimate the true accuracy of learned classifiers. This is an important question for any autonomous learning system that must es- timate its accuracy without supervision, and also when classifiers trained from one data distribu- tion must be applied to a new distribution (e.g., document classifiers trained on one text corpus are to be applied to a second corpus). We first show how to estimate error rates exactly from unlabeled data when given a collection of com- peting classifiers that make independent errors, based on the agreement rates between subsets of these classifiers. We further show that even when the competing classifiers do not make indepen- dent errors, both their accuracies and error de- pendencies can be estimated by making certain relaxed assumptions. Experiments on two data real-world data sets produce estimates within a few percent of the true accuracy, using solely un- labeled data. These results are of practical signif- icance in situations where labeled data is scarce and shed light on the more general question of how the consistency among multiple functions is related to their true accuracies.
Pages: 682-691
PS Link:
PDF Link: /papers/14/p682-platanios.pdf
AUTHOR = "Emmanouil Platanios and Avrim Blum and Tom Mitchell",
TITLE = "Estimating Accuracy from Unlabeled Data",
BOOKTITLE = "Proceedings of the Thirtieth Conference Annual Conference on Uncertainty in Artificial Intelligence (UAI-14)",
ADDRESS = "Corvallis, Oregon",
YEAR = "2014",
PAGES = "682--691"

hosted by DSL   •   site info   •   help