判定有限自动机可重置的二次多项式下界OA北大核心CSTPCD
On quadratic lower bounds for deciding resettable finite automata
对有限自动机重置判定问题和道路着色问题的时间复杂性下界进行研究,证明了这两个问题的条件下界分别是关于自动机状态数和有向图顶点数的二次多项式.这说明当指数时间猜想(ETH)成立时,两个问题现有的二次算法很大可能就是理论上最好的.对于有限自动机重置判定问题,建立从两个有限自动机交问题到它的细粒度归约,利用细粒度归约的传递性,得到其下界是关于状态数的二次多项式;对于道路着色问题,由于求解问题至少与验证问题的解一样难,可将判定某个有限自动机是否可重置看成验证道路着色问题的解,从而得出其下界是关于顶点数的二次多项式.
The time lower bounds of deciding resettable finite automata problem and road coloring problem were studied.It is proved that the lower bounds of these two problems are quadratic polynomials about the number of automata states and the number of vertices of directed graph,respectively.This shows that under exponential time hypothesis(ETH)the existing quadratic algorithms for both problems are almost the theoretically best.For the former problem,the fine-grained reduction from the intersection of two finite automata problem to it is established,and with transitivity of fine-grained reduction it is proved that the lower bound of determining whether the finite automaton is resettable or not is a quadratic polynomial about the number of the states.For the latter,because solving a certain problem is at least as hard as verifying one of its solutions,the proof regards deciding whether an automaton is resettable or not as verifying a solution of road coloring problem,and then it is proved that the lower bound of road coloring problem is a quadratic polynomial about the number of vertices.
朱凯;毛宜军;梁早清
华南农业大学数学与信息学院,广东 广州 510642华南农业大学数学与信息学院,广东 广州 510642||广东省食品质量安全重点实验室,广东 广州 510642
计算机与自动化
自动机重置判定问题道路着色问题细粒度复杂性指数时间猜想两个自动机交
deciding resettable finite automata problemroad coloring problemfine-grained complexityexponential time hypothesisintersection of two finite automata
《华中科技大学学报(自然科学版)》 2024 (002)
42-48 / 7
广东省科技计划资助项目(2020B1212060059);广东省农产品质量安全共性关键技术研发产业创新团队专题专家(2019-2023)项目(2019KJ130).
评论