B
Bianca Zadrozny
Researcher at IBM
Publications - 87
Citations - 6805
Bianca Zadrozny is an academic researcher from IBM. The author has contributed to research in topics: Feature selection & Binary classification. The author has an hindex of 26, co-authored 85 publications receiving 5921 citations. Previous affiliations of Bianca Zadrozny include Federal Fluminense University & University of California, San Diego.
Papers
More filters
Proceedings ArticleDOI
Transforming classifier scores into accurate multiclass probability estimates
Bianca Zadrozny,Charles Elkan +1 more
TL;DR: This work shows how to obtain accurate probability estimates for multiclass problems by combining calibrated binary probability estimates, and proposes a new method for obtaining calibrated two-class probability estimates that can be applied to any classifier that produces a ranking of examples.
Proceedings ArticleDOI
Learning and evaluating classifiers under sample selection bias
TL;DR: This paper formalizes the sample selection bias problem in machine learning terms and study analytically and experimentally how a number of well-known classifier learning methods are affected by it.
Proceedings Article
Obtaining calibrated probability estimates from decision trees and naive Bayesian classifiers
Bianca Zadrozny,Charles Elkan +1 more
TL;DR: It is concluded that binning succeeds in significantly improving naive Bayesian probability estimates, while for improving decision tree probability estimates the recommend smoothing by -estimation and a new variant of pruning that is called curtailment.
Proceedings ArticleDOI
Cost-sensitive learning by cost-proportionate example weighting
TL;DR: Costing is proposed, a method based on cost-proportionate rejection sampling and ensemble aggregation, which achieves excellent predictive performance on two publicly available datasets, while drastically reducing the computation required by other methods.
Proceedings Article
Learning Character-level Representations for Part-of-Speech Tagging
TL;DR: A deep neural network is proposed that learns character-level representation of words and associate them with usual word representations to perform POS tagging and produces state-of-the-art POS taggers for two languages.