چکیده مقاله
The research establishes an original framework which combines Explainable Machine Learning XML with Earned Duration Management EDM metrics to achieve project control and precise project duration prediction The low success rate of projects with substantial funding and the opaque nature of machine learning models compared to traditional EDM static formulas create a clear need for a new approach The case study presents a three layered system which combines Monte Carlo simulation for uncertainty modeling with Gradient Boosting prediction and SHAP explainability The proposed method creates training data through 50,000 simulations to enable both forward looking proactive risk management and backward looking root cause identification of deviations The integration of EDM with simulation and XAI produces more accurate and interpretable predictions which increases project managers' trust in the system The research establishes a method to convert project control from its current reactive state into a proactive system
کلیدواژهها
نویسندگان
محمد فتاح
Depatment of Industrial Engineering, Iran University of Science and Technology, Tehran, Iran
سیامک حاجی یخچالی
Depatment of Industrial Engineering, University of Tehran, Tehran, Iran
Narges Ghobadi
Depatment of Industrial Engineering, Amirkabir University of Technology, Tehran, Iran
شیوه ارجاع
Fattah, Mohammad and Haji Yakhchali, Siamak and Ghobadi, Narges,1404,Intelligent Project Control 4.0: Integrating Explainable AI with Stochastic EDM,11th International Conference on Industrial and System Engineering,Mashhad
ارائهشده در
مجموعه مقالات یازدهمین کنفرانس بین المللی مهندسی صنایع و سیستم ها2 مهر 1404 · مشهد