当前位置: X-MOL 学术Phys. Rev. Appl. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Analog-Quantum Feature Mapping for Machine-Learning Applications
Physical Review Applied ( IF 4.6 ) Pub Date : 2020-09-14 , DOI: 10.1103/physrevapplied.14.034034
Moslem Noori , Seyed Shakib Vedaie , Inderpreet Singh , Daniel Crawford , Jaspreet S. Oberoi , Barry C. Sanders , Ehsan Zahedinejad

Quantum information processing is likely to have a far-reaching impact in the field of artificial intelligence. Noisy, intermediate-scale quantum devices provide a platform for exploring the possibility of attaining a quantum advantage through hybrid quantum-classical machine-learning algorithms. One example of such a hybrid algorithm is “quantum kitchen sinks,” which builds upon a classical algorithm known as “random kitchen sinks” to leverage a gate model quantum computer for machine-learning applications. We propose an alternative algorithm called “analog-quantum kitchen sinks” (AQKSs), which employs an analog-quantum computer for mapping data features into new features in a nonlinear manner. The new features can then be used by a classical algorithm to perform machine-learning tasks. We show the effectiveness of our algorithm for performing binary classification on both a synthetic dataset and a real-world dataset by simulating the operations of a quantum annealer. We demonstrate that the AQKS algorithm reduces the classification error of a linear classifier from 50% to 0.6% for the synthetic dataset and from 4.4% to 1.6% for the other dataset. Our proposed AQKS algorithm presents the possibility to use current quantum annealers for solving practical machine-learning problems.

中文翻译:

机器学习应用的模拟量特征映射

量子信息处理可能会在人工智能领域产生深远的影响。嘈杂的中级量子设备提供了一个平台,用于探索通过混合量子经典机器学习算法获得量子优势的可能性。这种混合算法的一个示例是“量子厨房水槽”,它基于一种称为“随机厨房水槽”的经典算法,可以将门模型量子计算机用于机器学习应用。我们提出了一种称为“模拟量子厨房水槽”(AQKSs)的替代算法,该算法使用模拟量子计算机以非线性方式将数据特征映射到新特征。然后,经典算法可以使用这些新功能来执行机器学习任务。我们通过模拟量子退火炉的操作,展示了我们的算法在合成数据集和现实数据集上执行二进制分类的有效性。我们证明了AQKS算法减少了线性分类器的分类误差500.6 用于合成数据集,并来自 4.41.6对于其他数据集。我们提出的AQKS算法提出了使用当前的量子退火器解决实际的机器学习问题的可能性。
更新日期:2020-09-14
down
wechat
bug