The goal of this work is to improve the performance of differential protection schemes during false data injections (FDIs) by introducing multiple linear regression (MLR)-based cyber detection logic. Using the current information from the sending and receiving ends of the transmission line, the MLR logic delivers the FDI state. To find errors and to prevent malfunctions during cyberattacks, the differential protection method integrates logic. Furthermore, threshold setting is provided for the MLR logic to give appropriate decisions based on the training data. Performance assessment of the method is carried out with the new set of data to validate the applicability.