Discounted Cost Register Automata (DCRA) associate costs with strings in a regular manner using the operation of discounted sum. The min-cost optimization problem for DCRAs corresponds to computing shortest paths in graphs with more general forms of discounting than the well-studied notion of future discounting. We present solutions to two classes of such shortest path problems: in presence of both past and future discounting, we show the decision problem is NP-complete, but has a polynomial-time approximation scheme; in presence of two future discounting criteria that are composed in a prioritized manner, we show that the problem is solvable in Nexptime. © 2013 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Alur, R., Kannan, S., Tian, K., & Yuan, Y. (2013). On the complexity of shortest path problems on discounted cost graphs. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7810 LNCS, pp. 44–55). Springer Verlag. https://doi.org/10.1007/978-3-642-37064-9_6
Mendeley helps you to discover research relevant for your work.