Состоялся курс лекций профессоров М. Грабиша и А. Рузиновска-Грабиш
Международная научно-учебная лаборатория анализа и выбора решений организовала визит профессоров Мишеля Грабиш и Агнешки Рузиновска-Грабиш (University Paris I, Centre d’Economie de la Sorbonne). 24 ноября 2017 года состоялся курс лекций для всех желающих.
(1) “Multicriteria decision making with interactive criteria”
We make a general introductrion to multicriteria decision problems in the framework of conjoint measurement and multiattribute utility theory, and explain the central notion of mutual preferential independence and weak independence. We show how capacities (nonadditive measures) arise naturally and permit to model interaction between criteria. Two extensions of capacities are the Choquet integral model and the multilinear model. We explain under which conditions on conjoint measurement these models can live. Finally, we focus on the notion of interaction and its different definitions, and we show how they are related to the previous models.
(2) “Dynamic models of strategic network formation”
The traditional literature on network formation is basically divided into two branches: one that uses a random network approach (a link is formed by pure chance), and one that models strategic network formation (a link is formed by strategic interaction). In the random networks approach, one usually focuses on dynamic models and examines if emerging networks exhibit some real-world network features. The literature on strategic network formation assumes that agents form links by maximizing their utility functions, and focuses on the analysis of stability and efficiency. Although many models of strategic network formation are developed in static frameworks, there exist a branch of dynamic models of network formation with strategic interactions. More recently, also network formation has been combined with games on networks whose analysis focuses on the impact of the network structure on individual decisions. This recent literature studies the simultaneous determination of actions and links. In the present talk, we provide an overview of important contributions that apply a dynamic approach to strategic network formation.