In this work, we present an automatic framework for hyperparameter selection in Convolutional Neural Networks. In order to achieve fast evaluation of several hyperparameter combinations, prediction of learning curves using non-parametric regression models is applied. Considering that “trend” is the most important feature in any learning curve, our prediction method is focused on trend detection. Results show that our forecasting method is able to catch a complete behavior of future iterations in the learning process.
CITATION STYLE
Cardona-Escobar, A. F., Giraldo-Forero, A. F., Castro-Ospina, A. E., & Jaramillo-Garzón, J. A. (2018). Efficient hyperparameter optimization in convolutional neural networks by learning curves prediction. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10657 LNCS, pp. 143–151). Springer Verlag. https://doi.org/10.1007/978-3-319-75193-1_18
Mendeley helps you to discover research relevant for your work.