A New Forecasting Model for Container Throughput Based on Fuzzy Evidential Reasoning

Abstract

Integrating experts’ judgment with mathematical model outputs proves to be an effective option for the performance improvement of container throughput forecasting in a volatile economic environment. However, judgments, even on the same issue, differ a lot owing to experts’ various backgrounds. Besides, expert judgment frequently ambiguously presents itself. To tackle the above problem, this paper proposes a fuzzy-evidential-reasoning-based forecasting model (FERFM), which uses the fuzzy set theory to represent expert judgment’s ambiguity, and the evidential reasoning theory to integrate the opinions of different experts. For validation purposes, this paper compares FERFM with two widely used models (ARIMA and ANN) in terms of their forecasting performance based on Qingdao Port container throughput data. The results clearly show the superiority of the FERFM over ARIMA and ANN model, which indicates that FERFM could be a new effective container throughput forecasting tool in a volatile economic environment.