Neuro-Fuzzy Methods of Forecasting Quantitative Indicators of Logistic Activity

This article discusses the container transportation market in the Leningrad region and the need for new predictive models using neural fuzzy networks. Different methods of structural and parametric identification, such as ANFIS, R-ANFIS, Fuzzy-Partitions, SCRG, GD, LSE, PSO, ABC and FA are considered and their advantages and disadvantages are pointed out. It is proposed to adapt these methods to create predictive neural fuzzy models that can be used to assess the quantitative performance of logistics activities. The authors also emphasize the practical importance of accurate predictive models for determining key indicators in transport and logistics activities, and point out the limitations in the ability to reliably predict certain factors under resource and time constraints.