Forecasting The Air Temperature at a Weather Station Using Deep Neural Networks
Debneil Saha Roy
Abstract
Air temperature forecasting is an interesting research topic as changes in it affect our day to day lives. This type of forecasting is well suited for analysis by deep learning techniques. With the wide availability of weather observation data nowadays, these approaches can be utilized effectively. This work explores the application of deep learning models to air temperature forecasting in order to accurately predict it over two forecast horizons. Three deep neural networks are used in this study, namely, Multi-Layer Perceptron (MLP), Long Short Term Memory Network (LSTM) and a combination of Convolutional Neural Network (CNN) and LSTM. The predictive performance of these models is compared using two evaluation metrics. The results show that the combination of convolutional neural network and LSTM outperforms the other models in both the forecast horizons.