Keisuke Bandobando Lab.

Associate ProfessorKeisuke Bando

Overview of Research

This laboratory studies how to decide a matching between agents from two disjoint sets such as men and women and firms and workers. We analyze this problem by using game-theoretic modeling and analysis. The purpose of our study is to propose a desirable matching mechanism (or algorithm) that is applicable to real-life matching problems.

Major Taught Courses

Laboratories in industrial and systems engineering (6), Seminar in industrial and systems engineering, Bachelor’s thesis, Probability, Economic system and modeling, etc.

Research Contents

I’m interested in analyzing social situations using game theory and microeconomics. My specific research topics are given as follows.

①Matching problem

There are many situations where a partnership (or matching) is formed between people (or organizations or objects). In such cases, a matchmaker often decides who matches with whom based on the reported preferences of participants seeking partners (Figure 1). Examples include assigning students to laboratories or medical interns to hospitals. In such situations, a matchmaker needs to design an efficient algorithm that finds a desirable matching for participants. We analyze this problem using game theory.

  • Figure 1: matching process

②Network matching problem

A classical matching problem analyzes a matching between people from two disjoint sets, such as men and women. On the other hand, there are many real-life problems with more complicated structures. For example, trading between buyers and sellers often involves an intermediary such as a real estate agent. Three or more firms may collaborate in producing new products or services. We analyze whether there exists a desirable matching (or agreement) for participants in such situations.

③Mechanism Design

As mentioned above, a matchmaker needs to collect information about participants’ preferences to decide a matching. Should the participants reveal their true preferences to a matchmaker? This is not always the case. This means that participants may have better partners by misreporting their preferences. As a result, an inefficient matching may be achieved. To avoid this problem, we need to design matching algorithms by considering the strategic behavior of participants, which is called a mechanism design problem.

Recent Research Theme

  • Dynamic matching problems
  • Matching problems with contracts

Educational Policies

I want students to acquire a good logical thinking ability through learning game theory. In addition, our research may involve programming and simulation as well as mathematics. I believe that such skills will be helpful for the future of students.