Title: Individual model selection
Speaker: Prince Osei, 杏吧原创 University
Location: HP 4351 (MacPhail Room) – 杏吧原创 University
Date: Friday,聽March 24,聽2023
Time: 1:00 鈥� 2:00 pm
础产蝉迟谤补肠迟:听 In the study of decision-making patterns by individuals, such as the Iowa gambling聽task (IGT) process, there are several competing models suitable for聽describing a聽participant or group of individuals. Selecting a single best overall model typically聽comes at the cost of misallocating some individuals for the聽sake of the population.聽The calibration of the models requires re-running these models several times whenever聽an individual or set of individuals are moved or reshuffled among聽these models.聽Usually, there is a fixed computational budget, and can only afford to run each model聽once. An individualized model selection strategy that聽uses the Bayesian information聽criterion (BIC) is proposed. The method is computationally efficient and easy to聽implement to allocate the subjects to these聽competing models. The proposed聽individualized model selection strategy is illustrated using a hand-picked sample from聽the IGT data set.