Authors: Javier Lopez-Piqueres, Jing Chen
Published on: May 15, 2024
Impact Score: 7.4
Arxiv code: Arxiv:2405.09005
Summary
- What is new: Introduction of a novel family of tensor networks called constrained matrix product states (MPS) that integrate arbitrary linear constraints into their structure.
- Why this is important: Existing tensor networks struggled to incorporate linear constraints effectively, limiting their application in combinatorial optimization problems.
- What the research proposes: The new constrained MPS with a quantum region concept allows for the integration of any linear constraints while maintaining efficiency in processing.
- Results: Demonstrated superior performance in solving the quadratic knapsack problem compared to a leading nonlinear integer programming solver.
Technical Details
Technological frameworks used: Constrained matrix product states (MPS) and quantum region fusion rules
Models used: Canonical forms for MPS, unsupervised training strategy
Data used: Quadratic knapsack problem datasets
Potential Impact
Markets and companies in optimization software, particularly those dealing with combinatorial optimization and integer programming
Want to implement this idea in a business?
We have generated a startup concept here: QuantOptiNet.
Leave a Reply