Over the past few years his research has focused on developing a methodology in Machine Learning called the "Weight-Shape Decomposition" of a probability function such as the Parzen Window distribution. This provided a new basis for Quantum Clustering (which he developed in 2001), and opened the possibility for novel methods of image analysis. He has applied the latter to MRI brain images of mammals.
Future directions include: to investigate how the Shape component can be incorporated as a predefined filter in layers of Deep Convolutional Networks; continued biomedical research in collaboration with Erez Persi, who is currently in NIH, with the potential of interesting biomedical applications.
