Title :
Immuno inspired approaches to model discrete time series at state space
Author :
Giesbrecht, Mateus ; Bottura, Celso Pascoli
Author_Institution :
Machine, Components & Intell. Syst. Dept. (DMCSI), Campinas State Univ. (Unicamp), Campinas, Brazil
Abstract :
In this paper a new method for discrete time series state space modeling is proposed. The method is based on viewing the modeling problem as a constrained optimization problem. To solve the constrained optimization problem three imuno-inspired algorithms are proposed. An example is proposed to compare algorithms performance. Although the developed algorithms are dedicated to an specific problem, some ideas proposed in this paper can be used to solve any constrained optimization problem with immuno inspired algorithms.
Keywords :
optimisation; time series; constrained optimization problem; discrete time series state space modeling; immuno inspired approach; Cloning; Covariance matrix; Equations; Mathematical model; Matrix decomposition; Optimization; Time series analysis;
Conference_Titel :
Advanced Computational Intelligence (IWACI), 2011 Fourth International Workshop on
Conference_Location :
Wuhan
Print_ISBN :
978-1-61284-374-2
DOI :
10.1109/IWACI.2011.6160107