Time-series prediction problems have been effectively solved by deep neural networks lately given their ability to understand temporal characteristics found in time series. In this study, a deep learning-based flood occurrence prediction method is presented for the successful interpretation of weather events and meteorological data with higher accuracy. The proposed model is evaluated on the United States National Climate Data Center (NCDC) dataset, NCDC storm events. Correlation analysis was performed on the meteorological and weather phenomenal parameters for choosing the appropriate parameters. The experimental results show that the model achieves 87.8% accuracy while predicting floods in the United States from the year 2013 to 2019.
To improve the driving safety triggered by driver’s behavior recognition in an in-car environment, we propose to use depth cameras mounted in a car to generate behavior models generated by a deep learning algorithm for a driver’s behavior classification. The contribution of this paper is trifold: 1) The proposed multi-view driver behavior recognition system can handle the occlusion problem happened in one of the cameras; 2) Using the recurrent neural network can effectively recognize the continuous time behavior; 3) the average recognition accuracy of proposed systems can achieve 83% and 88%, respectively.