We have several research activities as listed below:
We develop machine learning algorithms on hardware (“bare-metal”) such as, FPGA accelerator boards, and massively parallel machines such as the DGX station.
Examples of applications: image classification using universal image distance, data compression, and large-width learning algorithm.
We study the statistical and mathematical theory of machine learning. This involves analyzing the complexity of learning of different problems, such as learning pattern classification, on various kinds of input spaces. For instance, on non-Euclidean spaces (which are also called distance spaces).
The following links point to some of our selected publications in algorithm design and development.
Most of the papers in the following link are on theory.
Joel Ratsaby, Ph.D.
Algorithms Research & Development Lab
Department of Electrical and Electronics Engineering
Ariel University of Samaria