Meta-learning Hyperparameter Performance Prediction with Neural Processes
Ying Wei 1 Peilin Zhao 2 Junzhou Huang 2
Abstract Free from the limitations of exhaustive search (Bergstra &
Bengio, 2012) being computationally expensive and model-
The surrogate that predicts the performance of specific methods (Keerthi et al., 2007) being too customiz-
hyperparameters has been a key component for able to ...


雷达卡




京公网安备 11010802022788号







