Unsupervised entropy-based selection of data sets for improved model fitting

@article{Ferreira2016UnsupervisedES,
  title={Unsupervised entropy-based selection of data sets for improved model fitting},
  author={Pedro M. Ferreira},
  journal={2016 International Joint Conference on Neural Networks (IJCNN)},
  year={2016},
  pages={3330-3337}
}
A method based on the information theory concept of entropy is presented for selecting subsets of data for offline model identification. By using entropy-based data selection instead of random equiprobable sampling before training models, significant improvements are achieved in parameter convergence, accuracy and generalisation ability. Furthermore, model evaluation metrics exhibit less variance, therefore allowing faster convergence when multiple modelling trials have to be executed. These… CONTINUE READING