EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

3746. Grey systems theory and a Markov chain model MCGM (1,1) applied to demand forecast for a 3PL provider

Invited abstract in session MD-25: Discrete, continuous or stochastic optimization and control in networks, transportation and design IV, stream Combinatorial Optimization.

Monday, 14:30-16:00
Room: 011 (building: 208)

Authors (first author is the speaker)

1. Francisco Trejo
Engineering, Universidad Anahuac
2. Rafael Torres Escobar
Facultad de Ingeniería, Universidad Anáhuac Mexico

Abstract

Post-pandemic logistics showed that just-in-time was not always the most successful supply alternative. The effect in the supply chain was a disruption, causing scarcity of material, delays in deliveries, and capacity constraints in transportation. This paper proposes to answer whether the amount of data or high level of uncertainty exist. The ability of models created with large amounts of data may be not necessarily the answer. The proposal is a novel alternative forecast model when the amount of data is not available (< 4 records) and integrates the past demand as a Markov Chain and offers a model's characterization to discriminate a set of data that is suitable for the Grey Systems Theory and a Markov Chain Grey Model MCGM (1,1). The model: 1) incorporates the past demand behavior, 2) establish the transition probability matrix, 3) defines enhanced forecast model through the GST GM (1,1), 4) develops the forecast, 5) measure its performance. The performance measurement was MAPE. The results exceed other models and methodologies such as: exponential smoothing, moving average and linear regression. The MAPE with the traditional methods was 6.58% vs. the MCGM (1,1) was 2.28% and provided a forecast range comparable to ARIMA.

Keywords

Status: accepted

Back to the list of papers