In our prior work , a concept of QoT was proposed to represent the confidence of the decisions and commands generated by the teleoperator. It reflected the operation status of the teleoperator. The investigation proved that the QoT varied over different times of a day due to the fatigue level. However, it did not consider the variation of the QoT during the same time period. In fact, the QoT of these teleoperators changed as they repeated the same task a few times. One main reason was because the QoT was also influenced by the training of the teleoperator. This paper investigated the effect of the training on the QoT. A B-Spline neural network based method was used to identify the QoT variation due to different training level online. Three experiment results were provided to investigate the influence of training on the QoT. The results were obtained in the end.