@inproceedings{9361a5045c914564ab1548a8d4158aaf,
title = "On the initialization of long short-term memory networks",
abstract = "Weight initialization is important for faster convergence and stability of deep neural networks training. In this paper, a robust initialization method is developed to address the training instability in long short-term memory (LSTM) networks. It is based on a normalized random initialization of the network weights that aims at preserving the variance of the network input and output in the same range. The method is applied to standard LSTMs for univariate time series regression and to LSTMs robust to missing values for multivariate disease progression modeling. The results show that in all cases, the proposed initialization method outperforms the state-of-the-art initialization techniques in terms of training convergence and generalization performance of the obtained solution.",
keywords = "Deep neural networks, Disease progression modeling, Initialization, Long short-term memory, Time series regression",
author = "{Mehdipour Ghazi}, Mostafa and Mads Nielsen and Akshay Pai and Marc Modat and Cardoso, {M. Jorge} and S{\'e}bastien Ourselin and Lauge S{\o}rensen",
year = "2019",
doi = "10.1007/978-3-030-36708-4_23",
language = "English",
isbn = "9783030367077",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer VS",
pages = "275--286",
editor = "Tom Gedeon and Wong, {Kok Wai} and Minho Lee",
booktitle = "Neural Information Processing - 26th International Conference, ICONIP 2019, Proceedings",
note = "26th International Conference on Neural Information Processing, ICONIP 2019 ; Conference date: 12-12-2019 Through 15-12-2019",
}