Boosting for probability estimation & cost-sensitive learning

ΟΜΙΛΗΤΗΣ:  Νικόλαος Νικολάου, University of Manchester, U.K.

ΤΙΤΛΟΣ:   Discovering the Brain through Big Data Exploration

ΧΡΟΝΟΣ:   Τετάρτη, 26/4/2017, 12:00 μμ

ΑΙΘΟΥΣΑ:  Αίθουσα Συνεδριάσεων Α56 (1ος όροφος)

ΠΕΡΙΛΗΨΗ:

Abstract: We provide a unifying perspective for two decades of work on cost-sensitive Boosting algorithms. We critique the relevant literature -consisting of more than 15 variants of the original algorithm- using four theoretical frameworks: Bayesian decision theory, functional gradient descent, margin theory, and probabilistic modelling. We find that only 3 of the published Adaboost variants are consistent with the rules of all the frameworks —and even they require their outputs to be calibrated to achieve this. Experiments on 18 datasets across 21 degrees of imbalance support the hypothesis —showing that once calibrated, they perform equivalently, and outperform all others. Our final recommendation —based on simplicity, flexibility and performance— is to use the original Adaboost algorithm with a shifted decision threshold and calibrated probability estimates. We then move on to the online setting which imposes the additional complication of having to decide whether to use new datapoints to update the parameters of the ensemble or those of the calibrator function. We propose resolving this decision with the aid of bandit optimization algorithms and present initial results suggesting superior performance to uncalibrated and naively-calibrated online boosting ensembles.