Prediction Of Anomalous Events With Data Augmentation And Hybrid Deep Learning Approach