High-Speed Remote Power Measurement by Communication of the Maximum and Minimum Measurement Value

A Semi-supervised Sparse Representation Neural Network for Error Estimation of Electricity Meters with Insufficient Tagged Samples

“…Three metrics, namely, the mean absolute error (MAE), the mean squared error (MSE) and the root mean square error (RMSE), were selected to evaluate the prediction results from different angles. The MAE is the most robust metric of the deviation between the real value and the predicted value [24], especially for large errors. The MSE and the RMSE can effectively characterize the dispersion of each method [25].…”