Soft Computing has became particularly important since the 90’s, when there was an increasing in the interest on the modeling of more complex systems, which began to appear in areas such as biology, medicine, human sciences, management sciences, social sciences, engineering, ecology, whose problems frequently are intractable for conventional mathematical and computational models.
The Research Group on Soft Computing of FURG is led by the teacher Graçaliz Pereira Dimuro, and aims at the development of theories, models, techniques, methods and software in order to offer solutions that are tolerant to the subjectivity, imprecision, incomplete, partial or conflicting information, uncertainty, partial truth and partial possibility, which appear in the modeling of complex systems and/or systems of imperfect information. The objective is to search for approximations or combinations of theories, models, techniques and methods that are capable of reflecting adequately the imperfection, reaching tractability, reliability and solutions of reasonable costs.
The research lines of the group are around the following themes: Fuzzy and Interval Computations (Interval mathematics, fuzzy sets and logic, interval fuzzy sets and logic, intuitionistic fuzzy sets and logic); Probabilistic Computation (Probabilistic models, bayesian nets, Markov Chains, Hidden Markov Models, Markov Decision Process, Markov Games, Monte Carlo simulation, Kalman filters); Bio-inspired Computation (Evolutionary systems, neural nets); Statistic Computation (Statistic inference, process statistic control, experiment design, performance analysis, reliability, data mining); Unconventional Computation (Models based on Domain Theory, models of interactive computation); Strategic Computation (Models based on Game Theory); Optimization; Hybrid Soft Models (Combination of soft models); Applications of soft modes (Applications to agents and multiagente systems, robotics, computational biology, social simulation, environment simulation, computational vision etc).
More information: Group's Mirror