Fig. 4: Prediction accuracy of deep learning model and the master equation on the noisy test set. | Nature Communications

Fig. 4: Prediction accuracy of deep learning model and the master equation on the noisy test set.

From: Deep learning enhanced Rydberg multifrequency microwave recognition

Fig. 4

The noise is white noise with mean μ = 0 and a standard deviation σ. The noise is added quantitatively using different σ values. Before and after addition of the noise, the data are scaled between 0 and 1 using their maximum and minimum. The deep learning model is trained in a training set without additional noise. In addition, we do not involve noise spectrum when solving the master equation. Each point is obtained after averaging of five predictions on the noisy data.

Back to article page