Cloning: a novel method for interactive parallel simulation
Proceedings of the 29th conference on Winter simulation
Fluent Learning: Elucidating the Structure of Episodes
IDA '01 Proceedings of the 4th International Conference on Advances in Intelligent Data Analysis
Hi-index | 0.00 |
As part of the DARPA Deep Green efforts, SAIC developed a multi-threaded and resolution approach to constructing and evaluating simulated futures to address the SimPath component. By making use of heuristically derived breakpoints in a provided plan, we can construct a series of possible futures that are grouped into what is referred to as the probabilistic fluent graph (PFG). Evaluation of the PFG is done through a multiple resolution and fidelity modeling system that applies computational resources to the areas of highest likelihood. In this paper, we will describe the architecture, processing, and data structures needed to bring this concept to fruition.