Project Management Using Self-Organizing Maps
Jamshid Parvizian, Hamed Tarkesh, Arezoo Atighehchian, Sara Farid · 2006
Abstract. Project management (PM) deals normally with different information variables that form a dynamic and high dimensional information space. Although better understanding from this space leads to more intelligent PM process, traditional PM methods cannot capture the complexity and dynamic nature of project information space. One way to overcome this problem is integrating system dynamics (SD) concept and traditional PM tools for providing better vision of dynamic behavior of project but some problems remain unsolved. For example, hidden information, the information lie among the relationships of accessible project information and SD's level variables, are not recognized. Solving this problem is the main motivation of this research. The main idea is integrating Self-Organizing Maps (SOM) and SD to map project's high dimensional information space into a two- or three-dimensional one. This decreases the complexity and leads to reveal more information about project condition and trend. In this way, many potential risks and disturbing events may be identified and managed easily. The proposed method is not only a PM analysis tool but also a unique reporting system that makes it possible to visualize the project condition.