Takuro Yamashita (Toulouse School of Economics) on "Splitting problem applied to Bayesian persuasion and mechanism design with externalities"

Speaker:  Takuro Yamashita - Toulouse School of Economics
  Wednesday, April 11, 2018 at 12:30 PM - Polo Santa Marta, Via Cantarane 24, Sala Vaona
A splitting problem is a problem of dividing possibly heterogenous objects into different groups. For example, (i) a seller of an object, say a used car with uncertain quality, may provide partial or noisy evidence about its quality (whereby cars with different qualities may be ``grouped together'') in order to maximize his expected profit; or (ii) a public authority may divide a mass of students with different ability levels into a number of schools in order to control the schools' peer effects and maximize their total attainment. The optimal splitting is characterized under certain conditions, and its ``natural'' implementation by monotone policies is presented.

Contact person
Angelo Zago

Publication date
November 16, 2017