Deep Learning is a machine learning method based on neural network architectures with multiple la... more Deep Learning is a machine learning method based on neural network architectures with multiple layers of processing units. It has been successfully applied to a number of data mining problems, particularly in the areas of image recognition and natural language processing. Deep learning for anomaly detection is an active and ongoing research topic. Supervised anomaly detection methods emphasise learning feature representations from known anomalies types. However, supervised approaches do not scale well to large data sets and may be unable to generalise well to unknown anomaly types. To counteract this, unsupervised learning techniques are often implemented which operate without labelled anomaly data, such unsupervised approaches do not utilise the sparse amount of normal data which is often easily available. This paper introduces a novel anomaly detection framework to address these problems. Instead of learning anomaly representations from known anomaly types or from unsupervised mod...
Dr Shoumen Datta, Research Scientist, Engineering Systems Division, Department of Civil and Envir... more Dr Shoumen Datta, Research Scientist, Engineering Systems Division, Department of Civil and Environmental Engineering; Co-Founder and Research Director, Forum for Supply Chain Innovation, School of Engineering, Massachusetts Institute of Technology, Room 1-179, ...
Uploads
Papers by Andrew Shields