D
Denis Deratani Mauá
Researcher at University of São Paulo
Publications - 84
Citations - 718
Denis Deratani Mauá is an academic researcher from University of São Paulo. The author has contributed to research in topics: Probabilistic logic & Inference. The author has an hindex of 15, co-authored 75 publications receiving 621 citations. Previous affiliations of Denis Deratani Mauá include Dalle Molle Institute for Artificial Intelligence Research.
Papers
More filters
Journal ArticleDOI
Evaluating credal classifiers by utility-discounted predictive accuracy
TL;DR: This paper derives, from a set of assumptions, a metric to evaluate the predictions of credal classifiers, which is related to the decision-maker's degree of risk aversion to the variability of predictions.
Proceedings Article
An ensemble of Bayesian networks for multilabel classification
TL;DR: A novel approach for multilabel classification based on an ensemble of Bayesian networks that assumes the features to be conditionally independent given the classes, thus generalizing the naive Bayes assumption to the multi-class case.
Journal ArticleDOI
Probabilistic inference in credal networks: new complexity results
TL;DR: It is shown that inferences under strong independence are NP-hard even in trees with binary variables except for a single ternary one, and it is proved that under epistemic irrelevance the polynomial-time complexity of inferences in credal trees is not likely to extend to more general models.
Journal ArticleDOI
Solving limited memory influence diagrams
TL;DR: A new algorithm for exactly solving decision making problems represented as influence diagrams that does not require the usual assumptions of no forgetting and regularity and is empirically shown to outperform a state-of-the-art algorithm on randomly generated problems of up to 150 variables and 1064 solutions.
Posted Content
Advances in Learning Bayesian Networks of Bounded Treewidth
TL;DR: This work presents novel algorithms for learning Bayesian networks of bounded treewidth that are empirically compared to each other and to state-of-the-art methods on a collection of public data sets with up to 100 variables.