具有小样本估计量的序贯最小优化算法
Sequential minimum optimization algorithm with small sample size estimators
摘要：
序贯最小优化是一种机器学习全局搜索训练算法。它适用于成本函数对给定的其他参数的可调参数的函数依赖性可以便宜地确定时。这个假设是由已知的门建立的量子线路所满足的。我们将其应用到光子电路中，其中出现了额外的挑战：低频率的巧合事件降低了算法的速度。我们建议修改算法，使小样本量估计器足以成功运行机器学习任务。我们证明了修改后的算法的有效性，将其应用到一个量子光学分类器与数据重新上传。
英文摘要：
Sequential minimum optimization is a machine-learning global search training algorithm. It is applicable when the functional dependence of the cost function on a tunable parameter given the other parameters can be cheaply determined. This assumption is satisfied by quantum circuits built of known gates. We apply it to photonics circuits where the additional challenge appears: low frequency of coincidence events lowers the speed of the algorithm. We propose to modify the algorithm such that small sample size estimators are enough to successfully run the machine learning task. We demonstrate the effectiveness of the modified algorithm applying it to a quantum optics classifier with data reuploading.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
