Title :
SCDF: a learning search algorithm
Author :
Zhang, Wei ; Hong, Shenggui
Author_Institution :
Comput. Sci. Dept., Liaoning Univ., Shenyang, China
Abstract :
By employing the Machine Learning technique to solve the heuristic search problem, a new kind of search strategy-Learning Search, with algorithm SCDF being one of its implementation-is proposed. The characteristics of SCDF are discussed and one of the conclusions is that after sufficient number of problem-solving, the mean complexity of SCDF is O(poly(N)), where N is the distance from the start node to a goal and poly(N) is a polynomial function of N
Keywords :
computational complexity; learning (artificial intelligence); problem solving; search problems; O(poly(N)); heuristic search; learning search; mean complexity; problem-solving; Algorithm design and analysis; Artificial intelligence; Computer science; Machine learning; Machine learning algorithms; Polynomials; Problem-solving; Search problems; State estimation; State-space methods;
Conference_Titel :
Intelligent Processing Systems, 1997. ICIPS '97. 1997 IEEE International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-4253-4
DOI :
10.1109/ICIPS.1997.669184
