Document Type : Research Paper

Authors

• Roya Soltani ¹
• Ali Nobakhti ²

¹ Assistant Professor, Department of Industrial Engineering, Faculty of Engineering, Khatam University, Tehran, Iran

² MSc in industrial Engineering, Department of Industrial Engineering, Faculty of Engineering, Khatam University, Tehran, Iran

Abstract

The performance of a project is influenced by multiple factors, the variations of which can lead to delays in execution, cost overruns, and reduced quality. The competence of the project manager in applying both soft and hard skills plays a crucial role in the successful completion of projects in today’s volatile, uncertain, complex, and ambiguous (VUCA) environment. Therefore, the design and development of a data-driven recommender system that can replace subjective judgments in the process of selecting qualified project managers, and accurately propose and recommend suitable candidates, appears to be both essential and inevitable. The recommender system proposed in this paper is based on multilayer feedforward neural networks trained with the Levenberg–Marquardt backpropagation algorithm, while employing a genetic algorithm to optimize the structure and weights of the neural network. The results demonstrate that the optimized neural network–based recommender system can identify and recommend qualified project managers with high accuracy. Compared with the non-optimized neural network, it achieves improved performance by reducing the MSE (from 0.351 to 0.094) and RMSE (from 0.593 to 0.307). This reduction in error indicates that the genetic algorithm effectively fine-tunes the neural network parameters, thereby enhancing system performance and enabling more accurate identification of competent project managers.

Keywords

• Recommender system
• project manager competence,, neural network, genetic algorithm, Project oriented organizations

Main Subjects

• project management