Resumen

ASAN CABALLERO, Laritza; ROJAS DELGADO, Jairo  y  JIMENEZ MOYA, Gerdys E.. Time Series Prediction Algorithm for Air Traffic Forecasting based on artificial neural networks. RCCI [online]. 2022, vol.16, n.4, pp. 84-100.  Epub 01-Dic-2023. ISSN 2227-1899.

Being able to know the future behavior of the air traffic flow has become an essential element for the success and development of this industry, especially considering its constant growth. Time series forecasting is a highly exploited area nowadays. Although each series is different, the use of models based on the Box-Jenkins methodology and linear regression is common. On the other hand, the trend of recent years is towards the use of neural networks. Within this group, there are many variants and parameters to be used. It cannot be guaranteed that one model is better than another, this depends on the results that each one has with the data of the series and the knowledge of the researcher. In this work, the use of two variants of a predictive algorithm based on artificial neural networks and neural networks whist short and long-term memory is proposed. The objective is to determine which of these allow to obtain the best results in terms of precision and execution time of the training. The results obtained show that the training for the model based on artificial neural networks is the most accurate, using less time for network training. In this result it is evident that the simplest model may be the best option for the prediction process.

Palabras clave : artificial neural networks; prediction; preprocessing; time series; air traffic.

        · resumen en Español     · texto en Español     · Español ( pdf )