P‐64: Panel Performance Prediction Using Domain Knowledge‐Guided Deep Learning

A robust deep learning model for predicting the performance of display panel is proposed by integrating domain knowledge into the framework. The model combines a simple learning part, which captures domain‐specific principles, with a deep learning part that identifies complex pattern, enabling accurate predictions even for unseen inputs. Experimental results show that the proposed method achieves high accuracy on both trained and untrained datasets while significantly reducing simulation runtime. The proposed model is efficient and interpretable, enabling rapid evaluation of multiple design variations.