Weather Forecasting Using Multilayer Perceptron Technique

Shifanaaz Abdulsab Nadaf, Afrasama A. Harlapur, Fathma Shekh, Aleena A. Sayed, Shashank Hegde

2025

Abstract

The Multilayer Perceptron (MLP) in weather forecasting is used for regression tasks based on input features such as pressure values, temperature, pressure values, wind. This proposed work focuses on evaluating the efficiency of MLP for accurate time series pattern predictions. This study incorporates ERA5 hourly data on pressure levels from 1940 to the present and uses a Feedforward Neural Network(FNN) MLP architecture. In addition, techniques such as the Cosine Annealing Learning Rate Scheduler and Hyperparameter Tuning are employed to analyze temporal relationships, perform feature selection and ultimately improve model performance. Experiments conducted with MLPs demonstrate competitive accuracy with MSE 91.92, MAE 7.02 and R² 0.9985 compared to traditional forecasting models, highlighting MLPs as a valid method for meteorological applications.

Paper Citation

in Harvard Style

Nadaf S., Harlapur A., Shekh F., Sayed A. and Hegde S. (2025). Weather Forecasting Using Multilayer Perceptron Technique. In Proceedings of the 3rd International Conference on Futuristic Technology - Volume 3: INCOFT; ISBN 978-989-758-763-4, SciTePress, pages 360-367. DOI: 10.5220/0013616600004664

in Bibtex Style

@conference{incoft25,
author={Shifanaaz Abdulsab Nadaf and Afrasama A. Harlapur and Fathma Shekh and Aleena A. Sayed and Shashank Hegde},
title={Weather Forecasting Using Multilayer Perceptron Technique},
booktitle={Proceedings of the 3rd International Conference on Futuristic Technology - Volume 3: INCOFT},
year={2025},
pages={360-367},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013616600004664},
isbn={978-989-758-763-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 3rd International Conference on Futuristic Technology - Volume 3: INCOFT
TI - Weather Forecasting Using Multilayer Perceptron Technique
SN - 978-989-758-763-4
AU - Nadaf S.
AU - Harlapur A.
AU - Shekh F.
AU - Sayed A.
AU - Hegde S.
PY - 2025
SP - 360
EP - 367
DO - 10.5220/0013616600004664
PB - SciTePress