MCDM'15 - paper no. 9


 

Back to MCDM'15 contents
 

ON A SEQUENTIAL DECISION PROCESS WHERE OFFERS ARE DESCRIBED BY TWO TRAITS

David M. Ramsey

Abstract:

This article presents a model of searching for some resource, e.g. a job, whose value depends on two quantitative traits. The decision maker observes offers in a random order and must accept precisely one offer. Recall of previously observed offers is not possible. It is assumed that the value of an offer is a linear function of these two traits, which come from a bivariate normal distribution. We consider the following four strategy sets: i) the decision on whether to accept an offer is based purely on the first trait, ii) any decision is only made after observing both traits, iii) after observing the first trait, the decision maker can either immediately accept, immediately reject or observe the second trait and then decide, iv) after observing the first trait, the decision maker can either immediately reject or observe the second trait and then decide. The goal of the decision maker is to maximize his expected reward, where the reward is equal to the value of the offer selected minus the search costs. The optimal strategy from each of these four sets is derived. An example is given.

Keywords:

sequential decision process, job search problem, choice based on several traits.

Reference index:

David M. Ramsey, (2016), ON A SEQUENTIAL DECISION PROCESS WHERE OFFERS ARE DESCRIBED BY TWO TRAITS, Multiple Criteria Decision Making (10), pp. 141-154

Full text:

download