PRESENTING KOHONEN NEURAL NETWORK MODEL FOR DETERMINING THE CONTRIBUTION OF EACH FACTOR INVOLVED IN THE DELAYS (CASE STUDY: TEHRAN-SHOMAL HIGHWAY-IRAN)

With proper investigation of project delays, they can bemade as lesson learning. So, in this research, a methodhas been presented to oer network- based pattern foridentifying the reasons of delays. For this purpose, reasonsof delay in Tehran-Shomal highway are indicated ina table for each group involved in the mentioned projectand the main source of delays of the four factors involvedin the project: employer, contractor (delay of materialsand equipment,human resource, supplier delays,delay in transportation,warehouse that is the prerogativeof contractor), consultant,external factorswere identied,categorized. Then, according to theexperts' opinions,interviews conducted with agentsworking in projects, all items of delays have been scoredfrom 0 to 10 since 1994 to 2015 in two cases:, the possibilityof delays in the project of Tehran-Shomal,theintensity rate,the impact of delays. In the last section,for the rst time, delays in Kohonen self-organizingneural network have been evaluated using MATLAB inthe program of SOFM. Delays in research ndings thatare not distinguished in two ways are not distinct, andoverlapp each other.