Abstract

BOUZA HERRERA, Carlos N.; MORENO, Agustín Santiago  and  SAUTTO VALLEJO, José M.. On signal-decision and its modeling. suma neg. [online]. 2018, vol.9, n.19, pp.1-7. ISSN 2215-910X.  https://doi.org/10.14349/sumneg/2018.v9.n19.a1.

We propose a model for several alternative approaches to the prediction of the probabilities involved by processing the obtained signals sent by the players, which may be companies competing in the market. They set a minimum level of expected utility and are interested in the probability that this expectation will be satisfied. These approaches are analyzed using Monte Carlo experiments, developed in the framework of Game Theory and different statistical models to make predictions about the probability of success or to obtain a certain utility. The behavior of the different proposed strategies is established, considering that the errors follow a Normal, double Exponential and Cauchy distribution. A comparison is made of the results given by the algorithms programmed R and SAS, finding that R are better in terms of δ, however GLIMIX PRC MIXED of SAS has lower values of V2 in all cases.

Keywords : Research development; dimension of the utility; likelihood of success; prediction of the probability; GLIM.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )