Lightning Talk Presenter

Debanjan Konar

Debanjan Konar

Purdue University

About

Debanjan Konar earned his Ph.D. from the Indian Institute of Technology Delhi, New Delhi, India. Currently, Dr. Konar is working as a Fulbright-Nehru Postdoctoral Fellow at Purdue University. Prior to that, he was a Postdoctoral Research Scientist at Helmholtz Zentrum Dresden-Rossendorf (HZDR), Dresden, Germany. He also worked as a Lead Research Scientist (Quantum Computing) at BosonQ Psi Pvt. Ltd. (Bangalore), India. He was a Helmholtz Visiting Fellow (Quantum Computing) at the Steinbuch Center for Computing (SCC), Karlsruhe Institute of Technology (KIT) in Karlsruhe, Germany. His research interests include quantum machine learning, hybrid classical-quantum algorithms, quantum-inspired neural networks, etc. Dr Konar has authored several articles published in top-notch AI journals and conferences. He is a Fulbright Fellow, a senior member of IEEE, an ACM member, Euro-Science member. He has won many prestigious awards, like the Fulbright- Nehru Postdoctoral Research Fellowships for 2023–2025, the Helmholtz Visiting Researcher Grant, and many more.
Debanjan Konar, Ph.D.
Fulbright-Nehru Postdoctoral Fellow (Quantum Computer Sc.),
Purdue Quantum Science and Engineering Institute (PQSEI), and School of Industrial Engineering, Purdue University, West Lafayette, USA

Tensor Ring Optimized Quantum-Enhanced Tensor Neural Networks for Large-Scale Simulations

The current quantum devices in the Noisy Intermediate Scale Quantum (NISQ) era usually have limited resources, making the recent quantum computing platform for real-life applications, particularly deep learning applications, extremely challenging. Moreover, the simulation of complex systems on a quantum simulator grows exponentially harder in both the number of qubits and the circuit depth or the number of layers. In this talk, I will introduce our proposed supervised Tensor-Ring Optimized Quantum- Enhanced Tensor Neural Networks (QTens-Net) to reduce feature space relying on the tensorization of classical features and to produce a quantum neural network (QNN) simulator that can surpass classical neural networks in the future generation for NISQ devices. In contrast to standard VQC that simulates QML algorithms, QTens-Net may provide a major advantage since it can represent the reduced feature space compared to classical counterparts in simulating large quantum systems. This characteristic is anticipated to generate a quantum benefit in post-quantum AI for large-scale applications.

Click here to view slides

Questions? Contact us!