SEMINAR:Revisiting Dynamic Dual Sourcing Problems with Data-centered...

Guest: Mustafa Hekimoğlu, Kadir Has University

Title: Revisiting Dynamic Dual Sourcing Problems with Data-centered Algorithmic Methods (IE, DSA)

Date/Time: 24 February 2025, 13:40

Location: FENS L045

Abstract: Companies use different criteria such as lead time, cost, and quality to evaluate suppliers; often using multiple suppliers with the aim of reducing stockout risk. However in many industries, there may be significant differences between the quality levels of different suppliers. This leads to a more difficult variant of the classical dual sourcing problem in the literature. In general, dual sourcing problems are difficult to solve as the conventional solution method, dynamic programming, suffers from the curse of dimensionality assuming Markovian property is satisfied. Recent developments in machine learning, data-centered optimization, artificial intelligence made it possible to tackle difficult dynamic control problems more algorithmically. Significant achievements are reported in academic and practical applications of these techniques to various supply chain problems. One common feature of these applications is that the solution technique is computationally expensive and, unlike theoretical results, it may suffer from severe deviations from the optimal (high variance).This seminar will review of dual sourcing problems, and their solutions with conventional methods (theory-induced heuristics) and novel, data-centered techniques (reinforcement learning).

Bio: Mustafa Hekimoğlu received his MSc and PhD degrees from Boğaziçi University and Erasmus University, respectively. In 2018 he joined the Department of Industrial Engineering at Kadir Has University, Istanbul. In 2024, he is appointed as an associate professor at Kadir Has University. Afterwards, he started his industrial career. Currently, he is working as a senior manager in an international IT company. His research focuses on stochastic processes, data science, inventory control, supply chain management, reliability and maintenance optimization.