Unit LeaderTakanori Maehara
Postdoctoral ResearcherTatsunori Taniai
We study the theory of discrete optimization. Discrete optimization problems are problems of finding the optimal solution from a finite number of candidates. Since many human decision making can be formulated in this form, solving discrete optimization problems is a fundamental technology in artificial intelligence. Ideally, discrete optimization problems can be solved by examining all candidates. However, when the number of candidates is huge due to the combinatorial explosion, it is impossible to examine all candidates in a realistic time. For such problems, we design an efficient algorithm with theoretical guarantee by using discrete convex analysis, graph theory, etc.