Leveraging Recurrent Neural Networks for Accurate Time Series Predictions

• Published in: 2023 International Conference on Emerging Research in Computational Science (ICERCS) pp. 1 - 5
• Main Authors: Jakhar, Sunil Kumar, Vennila, C., Rege, Pallavi R., Naval, Preeti, Haripriya, V, Vashisht, Nitish
• Format: Conference Proceeding
• Language: English
• Published: IEEE 07.12.2023
• Subjects: accompanied prediction; and mode; collection predictions; leverages; Performance evaluation; Prediction algorithms; Predictive models; Recurrent neural networks; Scientific computing; singular; Time series analysis; Training
• DOI: 10.1109/ICERCS57948.2023.10434021

this technical abstract gives a singular method to correct time collection prediction that leverages recurrent neural networks (RNNs). The technique starts off evolved off with pre-processing of enter statistics, accompanied by the design of a deep getting to know structure to generate and refine a series prediction. The elaborate patterns of a time series which might be commonly ignored by way of existing fashions are treated effectively via a recurrent neural network. These networks are applicable even in instances where the time series information is distinctly risky and unpredictable, as the version customizes its features consequently from the training datasets. The proposed approach is evaluated on a popular time series dataset, demonstrating stepped forward overall performance in comparison to present methods like device getting to know algorithms and deep learning models. The outcomes propose that recurrent neural networks are a promising solution for accurate time collection predictions.