Incorporating Domain Knowledge in Matching Problems via Harmonic Analysis
This work addresses a fundamental bottleneck in matching problems for machine learning and computer vision, though it is incremental as it builds on existing QAP methods with domain-specific adaptations.
The paper tackles the computationally hard quadratic assignment problem (QAP) in matching tasks by learning parameters from prior instances to accelerate solutions, showing that the new method can outperform existing approaches in practical domains.
Matching one set of objects to another is a ubiquitous task in machine learning and computer vision that often reduces to some form of the quadratic assignment problem (QAP). The QAP is known to be notoriously hard, both in theory and in practice. Here, we investigate if this difficulty can be mitigated when some additional piece of information is available: (a) that all QAP instances of interest come from the same application, and (b) the correct solution for a set of such QAP instances is given. We propose a new approach to accelerate the solution of QAPs based on learning parameters for a modified objective function from prior QAP instances. A key feature of our approach is that it takes advantage of the algebraic structure of permutations, in conjunction with special methods for optimizing functions over the symmetric group Sn in Fourier space. Experiments show that in practical domains the new method can outperform existing approaches.