English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Rainstorm prediction is of considerable importance for a wide range of applications, such as weather forecasting, disaster management, and flood monitoring. Predicting rare and extreme rainstorm events is challenging because only sparse historical samples are available for training. Additionally, rainstorm events are caused by many complex meteorological factors, involving heterogeneous meteorological observations. The interactions between these factors are also difficult to handle when making predictions. To address these challenges, we propose an integrated deep learning-driven prediction method based on adaptive attributed spatio-temporal affinities between spatio-temporal nodes, including spatio-temporal proximity and multi-dimensional meteorological attribute similarity using graph embedding. Based on the learned spatio-temporal affinity matrices, we apply graph convolutional networks to implement non-linear predictions. In particular, we develop an integrated loss function to address the class imbalance caused by rare rainstorm events. Our empirical evaluation results show that the proposed prediction method outperforms several competing state-of-the-art methods on two rainstorm datasets. We attribute the performance improvements of the proposed method to its ability to capture complex rainstorm development patterns using limited historical rainstorm samples and using heterogeneous spatio-temporal information.

Rainstorm prediction via a deep spatio-temporal-attributed affinity network