We are developing an anomaly detection system for video data that uses machine learning. The proposed system has two subsystems: feature extraction and anomaly detection. We developed two feature extraction systems. One uses traditional manual steps and the other uses machine learning, i.e., a neural network. For the anomaly detection system, we employ machine learning technology that we have developed for a cyber-attack detection system. Results confirm that both prototypes can detect anomalous events in experimental video data.