Add JIJ submissions for MIS#15
Conversation
|
Hi @ksk-jij , thank you for your submission. The CPU times seem very long. Can you describe your method in more detail please? Thank you! Best, |
|
Hi @mlschicker , Thank you for checking our submission. We've just updated the reference section accordingly. Please have a look at your convenience. Kind regards, |
| @@ -0,0 +1,2 @@ | |||
| Problem,Submitter,Affiliation,Date,Reference,Best Objective Value,Optimality Bound,Modeling Approach,# Decision Variables,# Binary Variables,# Integer Variables,# Continuous Variables,# Non-Zero Coefficients,Coefficients Type,Coefficients Range,Workflow,Algorithm Type,Paradigm,# Runs,# Feasible Runs,# Successful Runs,Success Threshold,Hardware Specifications,Total Runtime,Time to Solution,CPU Runtime,GPU Runtime,QPU Runtime,Other HW Runtime,Remarks | |||
| c-fat200-1,"Keisuke Sato, Hiromichi Matsuyama, Yuichiro Nakano, Kohji Nishimura, Yu Yamashiro","JIJ, JIJ, JIJ, JIJ, JIJ",2026-07-09,https://www.j-ij.com/. The algorithm builds on the SQOA framework (https://ieeexplore.ieee.org/document/11250318) with pre- and post-processing.,18,N/A,QUBO,200,200,0,0,1734,integer,-2 / 1,"JIJSQOASolver, which is under review for Qiskit Functions Catalog",Stochastic,Quantum and classical Hardware,10,10,1,0,"IBM Quantum backend(s): ibm_boston, Local CPU: MacBook Pro, Apple M4, 16GB.",7.49,N/A,2.49,0.0,5.0,N/A,Runtime is in seconds. Runtimes are averaged over runs. This solver is under the process to launch on Qiskit Functions Catalog. | |||
There was a problem hiding this comment.
| c-fat200-1,"Keisuke Sato, Hiromichi Matsuyama, Yuichiro Nakano, Kohji Nishimura, Yu Yamashiro","JIJ, JIJ, JIJ, JIJ, JIJ",2026-07-09,https://www.j-ij.com/. The algorithm builds on the SQOA framework (https://ieeexplore.ieee.org/document/11250318) with pre- and post-processing.,18,N/A,QUBO,200,200,0,0,1734,integer,-2 / 1,"JIJSQOASolver, which is under review for Qiskit Functions Catalog",Stochastic,Quantum and classical Hardware,10,10,1,0,"IBM Quantum backend(s): ibm_boston, Local CPU: MacBook Pro, Apple M4, 16GB.",7.49,N/A,2.49,0.0,5.0,N/A,Runtime is in seconds. Runtimes are averaged over runs. This solver is under the process to launch on Qiskit Functions Catalog. | |
| c-fat200-1,"Keisuke Sato, Hiromichi Matsuyama, Yuichiro Nakano, Kohji Nishimura, Yu Yamashiro","JIJ, JIJ, JIJ, JIJ, JIJ",2026-07-09,https://www.j-ij.com/. The algorithm builds on the SQOA framework (https://ieeexplore.ieee.org/document/11250318) with pre- and post-processing.,18,N/A,QUBO,200,200,0,0,1534,integer,-2 / 1,"JIJSQOASolver, which is under review for Qiskit Functions Catalog",Stochastic,Quantum and classical Hardware,10,10,1,0,"IBM Quantum backend(s): ibm_boston, Local CPU: MacBook Pro, Apple M4, 16GB.",7.49,N/A,2.49,0.0,5.0,N/A,Runtime is in seconds. Runtimes are averaged over runs. This solver is under the process to launch on Qiskit Functions Catalog. |
Spotted this discrepency while reviewing. This instance should have 1534 coefficients, corresponding to the number of edges if I am not mistaken.
There was a problem hiding this comment.
Perhaps this is not 1:1 correspondence.
There was a problem hiding this comment.
Hi @HuangJunye ,
what is written is correct in this case.
The number on non-zero coefficients in the basic QUBO formulation for MIS is exactly |V| + |E|.
(cf. https://github.com/ZIB-AOPT/QOBLIB/tree/main/07-independentset/models/binary_unconstrained)
In this specific case: 200 + 1534 = 1734.
So everything checks out.
Best,
Max
|
Hi @ksk-jij , the submission looks good! Can you confirm that you have used the standard QUBO formulation for MIS with P = 2? Best, |
|
Hi @mlschicker and @HuangJunye , Thank you for reviewing our submission! Yes, we used the standard QUBO formulation for MIS (the same as Model in https://github.com/ZIB-AOPT/QOBLIB/tree/main/07-independentset/models/binary_unconstrained) with the penalty coefficient 2. Kind regards, |
3 participants
Summary
Maximum Independent Set (MIS) submissions from JIJ Inc. for the 07-independentset benchmark.