國立政治大學應用物理所

本次由《謝憲毅 博士》演講交流，活動相關訊息如下，請參考。

演講主題：Continuous-variable quantum state tomography with machine learning
主講人：謝憲毅 博士後研究員 (清大光電所李瑞光教授實驗室)
演講時間：111年11月23日(三)14:10~15:30
地點：志希樓二樓 理學院會議室

※本場次演講採自由入場，歡迎有興趣的師生共同參與。

摘要：

Quantum state tomography is an technique which can completely reconstruct a quantum state from experimental data with numerical computation.  In this study, we developed a deep learning model with prior knowledge in laboratory.  It can be used to instantly infer the density matrix as well as the Wigner function of quantum states. With the help of this machine learning enhanced QST, the degradation of optical squeezed states induced by the environment can be extracted. By comparing with the traditional maximum likelihood(MLE) estimation, our benchmark also verify that it can infer quantum states with high fidelity by just using fewer data points even if the squeezing level is high. Such a fast and robust machine learning model open the possibility of feedback controlling quantum states.​