LAUSR

Abstract When creating an estimation model, determining which variables are efficient is of considerable importance. To strictly select efficient variables, it is necessary to define appropriate criteria for the task and to perform an exhaustive search, which is the search method that evaluates and compares all variable combinations. In this study, we apply an exhaustive search with a Gaussian process (ES-GP) to estimate coordination energy, which is related to performance of a Li-ion battery, and show that the estimation accuracy of ES-GP is significantly better than that of other methods in previous studies, such as MLR, LASSO and ES-LiR.