@article{Phooseekhiwe_Phoomvuthisarn_2022, title={Short Term Precipitation Forecasting using Recurrent Neural Networks, a Case Study of Suvarnabhumi Airport}, volume={12}, url={https://ph02.tci-thaijo.org/index.php/JIST/article/view/245615}, DOI={10.14456/jist.2022.2}, abstractNote={<p>Rainfall is one of the key important factors affecting human life. Accurate precipitation forecasting allows humans to better prepare for various activities that will happen in the future. However, in some situations, the availability of weather data is limited, making it difficult to make accurate forecasting. Currently, much research has chosen deep neural networks as an algorithm to train forecasting models. The main idea is to come up with relevant features at the architecture level. Based on this paradigm, it has been shown that the appropriate deep neural network architecture can flexibly mix and match features that are relevant in predictions. Consequently, most of the existing research then focuses on some of the techniques to improve the performance of the models without paying much attention on the issues of adding relevant features. However, when the training data is limited, deep neural networks might not scale very well, thus making it difficult to mix and match features for predictions. This imposes a research question of how effective the proposed forecasting models might be when not adding relevant features to the models. The aims of this research are to develop and compare the efficiency of various models for short-term precipitation forecasting with and without adding relevant features. The experiment consists of 2 parts: (1) The experiment by exploring whether the models with relevant featured provided can achieve higher accuracy compared to original models in equivalent environments, and (2) The experiment by comparing accuracy between 5 models (ARIMA, ARIMAX, RNN, LSTM and GRU). The weather dataset, which is very limited in quantity, used in this research was collected from Suvarnabhumi Airport. The results show that adding relevant features can enhance the forecasting performance of the model when the weather data is limited. In addition, the GRU model with relevant featured provided is the most effective prediction.</p>}, number={1}, journal={JOURNAL OF INFORMATION SCIENCE AND TECHNOLOGY}, author={Phooseekhiwe, Ragkana and Phoomvuthisarn, Suronapee}, year={2022}, month={Jun.}, pages={13–26} }