• Publications
  • Influence
Time for a change: a tutorial for comparing multiple classifiers through Bayesian analysis
TLDR
This work argues for abandonment of NHST by exposing its fallacies and, more importantly, offer better - more sound and useful - alternatives for it.
Should We Really Use Post-Hoc Tests Based on Mean-Ranks?
TLDR
The aim of this technical note is to discuss the inconsistencies of the mean-ranks post-hoc test with the goal of discouraging its use in machine learning as well as in medicine, psychology, etc.
Estimation of Constrained Parameters With Guaranteed MSE Improvement
TLDR
This approach has the parametrized structure of the maximum a posteriori probability (MAP) estimator with prior Gaussian distribution, whose mean and covariance parameters are suitably designed via a linear matrix inequality approach so as to guarantee, for any xisinX, an improvement of the mean-squared error (MSE) matrix over the least-squares (LS) estimators.
A Bayesian Wilcoxon signed-rank test based on the Dirichlet process
TLDR
This work proposes a nonparametric Bayesian version of the Wilcoxon signed-rank test using a Dirichlet process (DP) based prior, and addresses in two different ways the problem of how to choose the infinite dimensional parameter that characterizes the DP.
Statistical comparison of classifiers through Bayesian hierarchical modelling
TLDR
A Bayesian hierarchical model is proposed that jointly analyzes the cross-validation results obtained by two classifiers on multiple data sets and returns the posterior probability of the accuracies of the two classifier being practically equivalent or significantly different.
Robust Filtering Through Coherent Lower Previsions
TLDR
The classical filtering problem is re-examined to take into account imprecision in the knowledge about the probabilistic relationships involved, and a solution of the state estimation problem under such a framework that is very general and more robust to the presence of modelling errors in the system is derived.
Traffic intensity estimation via PHD filtering
TLDR
The paper will address the estimation of road traffic intensity from available measurements of mobile vehiclespsila coordinates by exploiting PHD (probability hypothesis density) filtering techniques based on the so called particle filter approach and road-map information.
Inference with Multinomial Data: Why to Weaken the Prior Strength
  • Cassio Polpo de Campos, A. Benavoli
  • Mathematics
    IJCAI
  • 16 July 2011
TLDR
It is argued that the strength of the Dirichlet prior should decrease at least k times faster than usual estimators do, because non-informative Bayesian estimators induce a region where they are dominant that quickly shrinks with the increase of k.
...
...