Past Event: Oden Institute Seminar

Advancements in Quantum Optimization Algorithms: From Efficient State Preparation to Improved Convergence

Linghua Zhu,

Abstract

In this talk, we will journey through a series of advancements in quantum optimization algorithms, particularly focusing on the Variational Quantum Eigensolver (VQE) framework and its derivatives. We begin by exploring the efficient preparation of multi-qubit trial states for VQE, emphasizing the significance of symmetries in determining optimal trial states. Transitioning to the Quantum Approximate Optimization Algorithm (QAOA), we introduce the Adaptive Derivative Assembled Problem Tailored - Quantum Approximate Optimization Algorithm (ADAPT-QAOA), a problem-tailored iterative version of QAOA that showcases faster convergence and reduced gate requirements. Delving deeper into the role of entanglement in quantum optimization, we highlight the flexibility of ADAPT-QAOA in entangling and disentangling qubits and the implications of entanglement entropy on algorithmic efficiency. We then discuss the preparation of Gibbs thermal states, vital for quantum applications, and present a novel approach that leverages low-depth circuits for state preparation. Lastly, we address the challenges in the VQE measurement step and propose strategies to enhance convergence and reduce shot requirements.

Biography

Linghua Zhu is a postdoctoral researcher at University of Washington. She received her PhD from New Jersey Institute of Technology in 2018. Her current research focuses on quantum machine learning and quantum-classical hybrid algorithms for molecular simulation, aimed at near-term and fault-tolerant quantum computing.