Predicting Rare Events through Deep Learning-based Anomaly Detection

• In Event: Machine Learning and Predictive Models

Brief Overview

This paper uses deep learning-based anomaly detection, a rapidly developing methodological approach in the field of deep learning, to overcome small sample size problems when predicting rare events such as armed conflicts and terrorist attacks.

Authors

• Budrul Ahsan, Philips Japan

• Sota Kato, International University of Japan

• Hirokazu Shimauchi, Tokyo Foundation for Policy Research