Abstract
The majority of today's weather forecasting studies have been focused on
complex physical models. These models are usually run on hundreds of nodes in a
High-Performance Computing system, which consumes a lot more power. Despite the
employment of these costly and complex tools, projections are frequently incorrect due
to inaccurate beginning conditions, measurements or a lack of understanding of
atmospheric dynamics. Furthermore, solving complex models like this often takes a
long time. The Internet of Things has helped any field that deals with technology.
Using an IoT device, a prototype based on a machine learning approach is proposed in
this study with an efficient framework, and implementation of an automated weather
prediction system based on Artificial Neural Network algorithms was designed and
developed. This system includes a technologically advanced irrigation system for our
convenience. Using ANN in this research, the weather for the next day appears to have
been predicted. The evaluation findings suggest that the model’s accuracy is sufficient
for existing works and their approaches.