The choice of prediction algorithm for the development of analytical software

Abstract

The choice of prediction algorithm for the development of analytical software

Kureichik V.M., Kartiev S.B.

Incoming article date: 09.12.2017

In this paper, a choice is made of the data processing algorithm necessary for the development of analytical software in the complex project ""Development and creation of high-tech production of an innovative system for the integrated accounting, recording and analysis of energy consumption and water consumption by industrial enterprises and utilities"" The review of existing algorithms and prediction methods in systems with a large number of parameters and a large epoch of analysis is made. A specific application for the desired algorithm is predicting the consumption of energy resources and water. Based on the review of algorithms, the algorithms chosen are most suitable for this task. The tandem use of the decision tree construction and the genetic prediction algorithm is considered. Further tasks that need to be solved for effective implementation of these algorithms in the development of analytical software are formulated.

Keywords: analytical software, forecasting, genetic algorithm