Solution strategy based on Gaussian mixture models and dispersion reduction for the capacitated centered clustering...

Area of Science:

• Operations Research
• Logistics and Supply Chain Management

Background:

• The Capacitated Centered Clustering Problem (CCCP) is crucial for optimizing industrial transportation and distribution.
• Solving CCCP is computationally complex, posing a significant challenge in logistics and supply chain management.

Purpose of the Study:

• To present a novel strategy for determining optimal facility locations in CCCP.
• To address the computational complexity associated with solving CCCP instances.

Main Methods:

• A strategy integrating Gaussian Mixture Models (GMMs) with dispersion reduction techniques was developed.
• The approach considers client point distribution patterns to identify likely facility locations.

Main Results:

• The proposed Dispersion Reduction GMMs approach achieved a mean error gap of less than 2.6% on large CCCP instances.
• This method demonstrated superior performance compared to Variable Neighborhood Search, Simulated Annealing, Genetic Algorithm, and CKMeans.
• The approach was faster in achieving results than Tabu-Search and Clustering Search.

Conclusions:

• The Dispersion Reduction GMMs strategy offers a computationally efficient and accurate solution for the CCCP.
• This method provides a competitive alternative for facility location optimization in logistics and supply chain management.