Paulo Cavalin  Paulo Cavalin photo         

contact information

Research Staff Member - Service Systems
Brazil Research Lab - Rio de Janeiro (RJ), Brazil



I am currently a Research Staff Member in the Social Data Analytics Group, at IBM Research - Brazil, conducting both theoretical and applied research in Pattern Recognition, Machine Learning, Computer Vision, and Service Sciences.

I received a Ph.D. degree in Automated Production Engineering/Computer Science at Ecole de Technologie Superieure (ETS) - Universite du Quebec, Montreal (QC) - Canada, in 2011, a M.Sc. degree in Applied Informatics at Pontificia Universidade Catolica do Parana (PUCPR), Curitiba (PR) - Brazil, in 2005, and a B.Sc. degree in Informatics at Universidade Estadual de Ponta Grossa (UEPG), Ponta Grossa (PR) - Brazil, in 2002. I received financial support from CAPES-Brazil for both the Ph.D. and the M.Sc., and from NSERC-Canada for the Ph.D..

Prior to joining IBM, I was a Professor at Universidade Federal do Tocantins (UFT), Palmas (TO) - Brazil. Previously, I have also been a Research and Teaching Assistant at ETS; a Visiting Researcher at Centre for Pattern Analysis and Machine Intelligence (CENPARMI) - Concordia University, Montreal (QC) - Canada; and a Researcher & Developer at Invisys - Sistemas de Visao Computacional, Curitiba.

Currently, my main research focus lies in investigating the adaptation of classifiers and the dynamic selection of the best members of a pool of classifiers to deal with pattern recognition problems subjected to a high-level of uncertainty, e.g. problems for which only a small training dataset is available at first or problems whose underlying concept drifts over time. I also have experience in developing systems for handwriting recognition, industrial inspection, and biometrics.