An Algorithm for Forecasting Future Trends

Abstract

The contemporary information landscape is characterised by a huge amount of data available for analysis using a variety of research tools and methods. Considering the limitations of using individual models and methods, it is worth employing an approach that combines functional and logical autoregression methods to conduct a more accurate analysis of trends and topics in the information space. Considering this context, this work aims to develop an algorithm to identify and analyse topics that would be relevant in the future using autoregression methods. The process begins with the quantification and normalisation of data, which significantly affect the quality of analysis. The main focus of this study is to implement the autoregression method to analyse long-term trends and predict future developments in the selected data. The proposed algorithm evaluates the forecast of these future developments and analyses graphical trends, thus conducting a more detailed study and modelling of future data dynamics. The regression coefficient is used as a quality criterion. The algorithm concludes with a polynomial function to help identify topics that will be relevant in the future. Overall, the proposed algorithm can be considered an effective tool for analysing and predicting future trends based on the analysis of historical data, thus contributing to the identification of prospects for technological development.

Keywords

• data quantification
• time series normalisation
• forecasting
• autoregression models
• data normalization
• trends