[17a-K306-1~10] FS.1 Focused Session "AI Electronics"

11:45 AM - 12:00 PM JST(2:45 AM - 3:00 AM UTC)

[17a-K306-10] Trial of audio anomaly detection using quantum machine learning -First trial for introducing to factory-

〇Takao Tomono¹, Kazuya Tsujimura² (1.Keio Univ., 2.TOPPAN HD)

Keywords:

audio,anomaly detection,quantum machine learning

In manufacturing sites such as factories, anomalies detection learning model are often made from one-class SVM that uses the coefficient parameters of an autoregressive (AR) model. In this study, we investigated the effect on the learning model by changing the classical kernel embedded into one-class SVM to a quantum kernel. We recorded different abnormal sound patterns from two different devices. For both anomalies, we confirmed that the accuracy would be improved by using the quantum kernel more than classical kernel.