Weather Prediction in Bangladesh Using Distinct Artificial Intelligence Techniques
No Thumbnail Available
Date
2024-05-30
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Weather forecasting holds immense significance in Bangladesh due to its heavy reliance on
agriculture and vulnerability to weather-related risks. This study aims to develop reliable
weather forecast models for Rainfall, Temperature, and Humidity using artificial intelligence
algorithms such as Decision Trees, Random Forests, K-Nearest Neighbors, Support Vector
Machines, and Multilayer Perceptron Neural Networks. Historical weather data from the
Bangladesh Meteorological Department (BMD) undergo preprocessing to handle missing
values and outliers. Various AI techniques are then applied, leveraging their ability to address
complex and nonlinear interactions in data. The prediction models are evaluated using
performance metrics like Mean Absolute Error and Root Mean Square Error, Root Mean
Square Error, R-square accuracy, R-square Value, and Relative Absolute Error. This research
contributes to advancing weather forecasting in Bangladesh, benefiting sectors like
agriculture, disaster management, and public safety. We divided our Temperature, Humidity,
and Rainfall Datasets into two intervals: 1999-2022 which consist of 34 weather stations and
1980- 2022 consist of 26 weather stations. And we used 80% of our data for training and 20%
for testing our models. We analyzed our data using both interval and did a detailed analysis
of the yearly, monthly, and seasonal trends of Temperature, Rainfall, and Humidity. It also
highlights the relationship between These weather parameters. From our results, we have
found that Random Forest and Extreme Gradient Boosting Perform better than the Support
vector machine, K-Nearest Neighbor, Multilayer neural network, and Decision Tree
algorithm. And It is also noticeable that The results of the time interval 1999- 2022 are better
than the time interval 1980-2022.
KEYWORDS: Artificial Intelligence; Machine Learning; Random Forest; Decision Tree;
Support Vector Machine; K-Nearest Neighbor; Extreme Gradient Boosting; MSE; RMSE;
MAE; RAE; R- Square Value; R-Square Accuracy.