ON THE APPLICATION OF MACHINE LEARNING AT MONITORING OF PARAMETERS OF LOGISTICAL SYSTEMS

P. 275–280.

Барашков А. В., Петровский Д. В.

This article discusses the creation and application of the machine learning approach for the dynamic control of the following parameters of logistic systems: the level of warehouse residues and the delay time of the supplier's machine when the goods are delivered to the company's warehouse. For the analysis of these parameters, two classifiers were created using the logistic regression method, which were able to classify the current state of the system and the level of threat in performing certain operations. For each parameter, a training and test sample were created, which were later used for training and testing classifiers. After completing of training, the monitoring system was able to classify new data in real time that had not been previously reported, which allowed to define the system state for an indefinite period of time. For the classifier, heuristically were created classification rules, which most accurately reflect the state of the logistics system

Keywords: logistic regression, logistics systems, classification, Wolfram Mathematica

In book

Логистика: современные тенденции развития: материалы XVII Междунар. науч.-практ. конф. 12, 13 апреля 2018 г.

Ч. 2: мат. докл. , СПб.: ГУМРФ имени адмирала С.О. Макарова, 2018.