US Patent 10984757 Machine learning method, machine learning system, and display system

To improve the display quality of a display device. To provide a method of correcting image data input to the display device. To provide a novel image correction method or an image correction system. Machine learning for a neural network correcting image data input to the display device is performed by the following method: second image data based on an image that is displayed on the display device by input of first image data to the display device is obtained; third image data is generated by obtaining a difference between the first image data and the second image data; fourth image data is generated by adding the first image data and the third image data; and a weight coefficient is updated so that output data obtained by input of the first image data to the neural network is close to the fourth image data.