A project portfolio selection model with project interaction & resources interdependency consideration using artificial neural networks

Abstract

A vast number of organizational projects as well as complexity of decision making process can cause particular challenges for project management and its leadership. In order to use organization assets and opportunities efficiently, it is necessary that manager implement a comprehensive multidimensional project portfolio management system that considers economic, social and technical details of the projects. Resource constraint compels managers to select operational proposal projects. Thus managers can maximize organizational utility due to project portfolio’s resource constraint. This study considers the interactive effects of project portfolio evaluation and sharing organizational project resources with respect to its evaluation and choice of the projects.. In this two-step model, first a branch and bound algorithm with resource interaction was utilized to screen maximal portfolio and, in the next step, each portfolio was evaluated based on artificial neural networks to rank the end project portfolios. Also, the ANN scores are strongly correlated with the DEA and COLS efficiency scores.

Keywords

• Artificial Neural Network
• interaction
• maximal portfolio
• Project portfolio selection
• resource interaction