Risk taking in stochastic PERT networks

Risk taking in stochastic PERT networks

0.00 Avg rating0 Votes
Article ID: iaor19961588
Country: Netherlands
Volume: 67
Issue: 2
Start Page Number: 221
End Page Number: 241
Publication Date: Jun 1993
Journal: European Journal of Operational Research
Authors:
Keywords: PERT
Abstract:

This paper studies a PERT problem which incorporates project managers’ risk taking behavior in stochastic environments. In this stochastic PERT network, which the paper refers to as RISKPERT, activity durations are nonnegative random variables, and a project manager uses a disutility function to evaluate each path and chooses that path which maximizes his expected disutility as an ‘optimal’ (or critical) path. This path, as opposed to the classical critical path, is used as a basis to analyze the PERT network and to estimate the probability of completion time. The general RISKPERT is difficult to solve; however, special cases when activity durations are statistically independent and disutility functions are linear, exponential, quadratic, or linear-exponential, are solvable exactly. The illustrative examples demonstrate that RISKPERT captures managers’ risk taking behavior and provides results that are more realistic than those of the classical PERT models. Furthermore, this paper extends some of the RISKPERT models to include multi-dimensional PERT networks.

Reviews

Required fields are marked *. Your email address will not be published.