Uma abordagem de apoio à decisão multicritério baseado em predição probabilística e em uma representação tensorial de dados
Betania Silva Carneiro Campello, Leonardo Tomazeli Duarte, João Romano

DOI: 10.14209/SBRT.2020.1570658231
Evento: XXXVIII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2020)
Keywords: Predição probabilística álgebra tensorial apoio à decisão incerteza
Abstract
Multiple criteria decision analysis aims at ranking alternatives by means of multiple criteria. In general, the data classically used are the data in only the instant of time in which the decision is made. However, these data are not representative when the consequences of the choice depend on the future data of the criteria. Thus, we propose to predict the criteria value, through their time-series structured as a tensor. We use a probabilistic prediction method, which provides the probability of the criteria taking specific values. In this approach, it is possible to quantify the uncertainty in decision making. Numerical experiments suggest that probabilistic prediction improves decision support.

Download