The Roychowdhury Group
Connecting Computer Science & Physics & Biology & Society
Professor Vwani Roychowdhury
Mining the connections between engineering, physics, biology, and society
using computational and information science
to shape our future
From the way we design quantum computers, to the size of our cell phones, to how we care for our children, the most exciting advances in our world emerge from the dynamic union of diverse bodies of knowledge, as captured through a common and interacting set of computational principles. These advances are both described by and enabled through the work I do to create interdisciplinary models of our reality.
My work has always tapped the energy generated by the intersection of diverse fields, both describing their synergy and catalyzing their transformation.
Indeed, in my view, every discipline is an information processing and computing system, driven by a succinct set of universal laws.
Hence, if one uses mathematical models to describe the foundations of different fields, the models can refine and inform one another. As models cultivate understanding of each discipline, they also create the ability to direct the discipline’s growth.
Search my site (including all published papers) here, or click below to explore by subject.
Areas of Interest
The Roychowdhury group has initiated a new area of study. We Explore The means by which to use publicly available trace data (from search engines & the internet) that is related to human behavior & perception. In effect, we discover stochastic models of the propagation of information, fame & sentiments in society. We employ analytical tools from statistical physics, Bayesian statistics & applied mathematics.
As an alternative to the classical computing paradigms, I got into the field of Quantum Computing and Information Processing in its infancy. Along with my colleague, Prof. Eli Yablonovitch we received multi-million dollar grants from DARPA and the ARO and formed an internationally recognized quantum-computing group at UCLA.
We looked at a number of fundamental issues in this field, including the capacity of Neural Networks to both learn and compute, and online learning algorithms using stochastic gradient algorithms. For example, we studied the role of depth, which is now a critical parameter in Deep Learning. We showed how depth plays a critical role in determining the size of a network, and also in ensuring the emergence of certain patterns.
In my continuing quest to find alternative models of computation and biological and nature inspired computing, I started exploring how robust and highly-adaptive emergent structures and functionalities appear in self-organized systems, and how to build engineering systems based on such principles. This led to pioneering work on modeling organic structures and the processes that led to the development of the World Wide Web (WWW), Peer-to- Peer (P2P) networks, and other emergent systems such as social networks, and online auctions.
