Finding similar weather patterns or analogs in historical datasets can be useful for data assimilation, forecasting, postprocessing and for attribution and impact studies. However, searching in those datasets is computationally expensive. A distance-regularized Siamese twin autoencoder (DIRESA) architecture is designed to preserve distance in latent space while capturing the nonlinearities in the datasets. The search can then be executed in latent space and the latent components thus obtained provide physical insight into the dominant modes of variability in the system. DIRESA is provided as an open-source Python package based on Tensorflow, using a CNN/Dense or custom encoder and decoder.
De Paepe, G & De Cruz, L 2024, 'DIRESA – A Deep Learning-based, nonlinear “PCA”', Large-Scale Deep Learning for the Earth System, 29/08/24 - 30/08/24.
De Paepe, G., & De Cruz, L. (2024). DIRESA – A Deep Learning-based, nonlinear “PCA”. Poster session presented at Large-Scale Deep Learning for the Earth System.
@conference{e195cd15fe59410d91ed260ca3b1a160,
title = "DIRESA – A Deep Learning-based, nonlinear “PCA”",
abstract = "Finding similar weather patterns or analogs in historical datasets can be useful for data assimilation, forecasting, postprocessing and for attribution and impact studies. However, searching in those datasets is computationally expensive. A distance-regularized Siamese twin autoencoder (DIRESA) architecture is designed to preserve distance in latent space while capturing the nonlinearities in the datasets. The search can then be executed in latent space and the latent components thus obtained provide physical insight into the dominant modes of variability in the system. DIRESA is provided as an open-source Python package based on Tensorflow, using a CNN/Dense or custom encoder and decoder.",
keywords = "Deep Learning, Dimension Reduction Methods, weather regimes",
author = "{De Paepe}, Geert and {De Cruz}, Lesley",
year = "2024",
month = aug,
language = "English",
note = "Large-Scale Deep Learning for the Earth System ; Conference date: 29-08-2024 Through 30-08-2024",
url = "https://cesoc.net/workshop-on-large-scale-deep-learning-for-the-earth-system/",
}