My research blog
This is an old version
I have had an exciting and diverse set of research experiences which I have outlined in my CV, of these, two of the most impactful projects for me were one in my fourth year, where I delved deeper into machine learning assisted physics research. I was fascinated by the workings of Neural networks and Random tree algorithms, which led me to take Data Science for my [Minor]; I soon went on to do my project with Dr. Nagaiah Chamakuri where we tried to find the [Quantum states that showed Superadditivity of Coherent Information when passed through a Noisy Quantum Channel]. This exposed me to a wide variety of scientific computing tools, meta-heuristic algorithms being one of them. The work I did here on metaheuristic algorithms, TensorFlow, and MPI parallelism helped me with my next project on using [Machine Learning for exploring non-locality in Quantum Networks without inputs] which was also my [Master’s thesis]. Network structures that obey locality constraints can be matched using Neural networks because their Direct Acyclic Graphs match, using this it has been shown that a sufficiently trained neural network reproducing a nonlocal distribution can be a good witness for its nonlocality. We expanded on the existing research to combinations of Bell states forming X states with two parameter entangled joint measurements and we have found the best measurement choices for these X states and their pattern. I am currently in the process of publishing my results which can be checked in my \href{https://ananthrishna.github.io/}{portfolio}.
Here understanding the results from the data driven models lead me to introspect the possible symmetries underlying the analytical calculations and led me to run better models to understand the problem, this method of leveraging the strengths of both analytical methodology and computational techniques suited me as it let me understand the problem from multiple angles giving it a comprehensive touch. I am also currently continuing this research to reframe the locality constraint of networks finally introduce a package for generic quantum network by consulting and studying few compatible approaches. Having found optimization here in the non-convex boundary of Quantum networks quite difficult along with dimensionality of the problem being quite high I was inspired to learn and come up with better hybrid optimizers to overcome this problem. The metaheuristic algorithms inspired from evolutionary schemes I learned during my Minor in Data Science helped here.
Taking inspiration from current research with reduced network output cardinalities we are trying to simplify and generalize the network’s nonlocalty nature. Understanding the results from the data-driven models led me to introspect the possible symmetries underlying the analytical calculations that are responsible and led me to run better models to understand the problem, this method of complimenting both analytical methodology and computational techniques let me understand the problem from multiple angles. The 64-dimensional target distribution being very close to the boundary of local and nonlocal distributions posed a great difficulty in the training of the neural network, the metaheuristic algorithms I learned from my earlier project were really handy here. Thus, I have been working on a hybrid optimizer that can overcome the dimensionality curse and also provide sufficient accuracy. The best measurement choices that have been found have also been checked for their noise robustness to ensure their applicability in real practical scenarios. I am currently in the works of publishing this result which can be found …
During spring 2021, I did a paper discussion as part of the Quantum Foundations course. under the guidance of Dr. Manik Banik (S.N.Bose NCBS) on the Bohr-Einstein debate and Bell’s theorem, It was a constructive group project where we read articles and taught each other. This was one of the projects that made me look forward to doing research in Quantum Information Theory. I found myself learning a lot better after learning as a part of a community and ever since I have looked forward to networking and exchanging ideas with like-minded peers. I am thankful to have done a research internship with Prof. Anil Shaji during the summer of 2019 in my second year. We worked on finding quantum bound states using variational principles particularly the Gauss-Seidel iterative algorithm for finding electron wave function and energy.
I have attended several workshops that are outlined in my CV; a particularly interesting one was Brain, Computation, and Learning (BCL) 2023 at IISc Bangalore. It was inspiring to meet a lot of expert researchers from varying fields and discuss with them the possibilities of research in an interdisciplinary between Quantum Information and Neuroscience. I have also attended external hackathons, including the Datathon IndoML 2023. I am also experienced in Image Processing and Curve Fitting using MatLab. I have designed websites and I am fairly experienced in Web development using NodeJS and Flash. I also have expertise in using Blender, Adobe Premiere Pro, Animation using Adobe Flash, Procreate, and Adobe Photoshop.