Nonlinear Prediction Models in Data Analysis

Authors

  • Željko Račić University of Banjaluka
  • Zoran Ž. Avramović PanEuropean University APEIRON Banjaluka, Bosnia and Herzegovina
  • Đuro Mikić PanEuropean University APEIRON Banjaluka, Bosnia and Herzegovina

DOI:

https://doi.org/10.7251/JIT2002106R

Abstract

The modern entrepreneurial sensibility of the company’s business implies directing the right information to the appropriate parts of the company at the right time. That is why it is necessary to digitalize processes as much as possible and make the organization “intelligent”, and its human resources, to the greatest extent, the knowledge workers. The application of neural networks, i.e. nonlinear prediction models, enables systematic analysis of data in the function of evaluating the behavior of the system. Neural networks are a powerful tool, especially for forecasting trends and forecasting based on historical data. The grouping method, i.e., the k-mean value algorithm, is used as a precursor to neural networks.

Author Biographies

Željko Račić, University of Banjaluka

University of Banjaluka

Zoran Ž. Avramović, PanEuropean University APEIRON Banjaluka, Bosnia and Herzegovina

PanEuropean University APEIRON Banjaluka

Đuro Mikić, PanEuropean University APEIRON Banjaluka, Bosnia and Herzegovina

PanEuropean University APEIRON Banjaluka

Downloads

Published

2021-09-20

Issue

Section

Чланци