Investment contexts are modeled as Partially Observable Markov Decision Processes (POMDP), processed by planning algorithms on a grid computing environment and simulated. Historical series are used to provide both the probability of changing from one state to another and the probability of being in a certain state when an event is reported by the sensors. These probabilities are used by an automated planning algorithm to create policies that try to maximize the profit. The execution of generated policies is simulated and evaluated.
Related Papers:
- Baffa, Augusto Cesar Espíndola; Ciarlini, Angelo E. M.; Modeling POMDPs for generating and simulating stock investment policies; Proceedings of the 2010 ACM Symposium on Applied Computing; Sierre, Switzerland
- Baffa, Augusto Cesar Espíndola; Ciarlini, Angelo E. M.;Planning under the Uncertainty of the Technical Analysis of Stock Markets; IBERAMIA 2010: 12th edition of the Ibero-American Conference on Artificial Intelligence; Bahía Blanca, Argentina, November 1-5,2010
