Performance of Deep Learning Algorithms to Predict the Monthly Rainfall Data of Rajshahi District, Bangladesh
DOI:
https://doi.org/10.3329/ijss.v25i1.81044Keywords:
Rainfall, Climatic variables, Machine Learning algorithms, deep learning algorithms, Rajshahi.Abstract
The economic development of Bangladesh highly depends on agriculture production and rainfall is one of the most influential factors. A number of variables, including temperature, relative humidity, wind direction, wind speed, and cloud cover, influence the likelihood of rainfall. There is currently a deficiency in the ability to accurately and precisely predict rainfall, which would be beneficial in a variety of industries, including flood prediction, water conservation, and agriculture. Recently, machine learning algorithms showed better performance for predicting climatic variables than tradition models. Using deep learning algorithms to forecast rainfall is an innovative method that makes use of sophisticated computer methods to examine complex patterns in meteorological data. So, in this paper we compare the forecasting performance of deep learning algorithms and machine learning algorithms in case of Rajshahi district in Bangladesh. The historical data from January 1964 to December 2017 is considered for study. The empirical results suggest that, for the subsequent timeframes, the deep learning algorithms MLP is the most suitable algorithm for forecasting the monthly rainfall data of this study area.
IJSS, Vol. 25(1), March, 2025, pp 39-54
151
40
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Department of Statistics, University of Rajshahi, Rajshahi

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.