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

Effective water resource management requires information on wateravailability, both in terms of quality and quantity, spatially and temporally.In this paper, we study the methodology behind Transfer Learning (TL) throughfine-tuning and parameter transferring for better generalization performance ofstreamflow prediction in data-sparse regions. We propose a standard recurrentneural network in the form of Long Short-Term Memory (LSTM) to fit on asufficiently large source domain dataset and repurpose the learned weights to asignificantly smaller, yet similar target domain datasets. We present amethodology to implement transfer learning approaches for spatiotemporalapplications by separating the spatial and temporal components of the model andtraining the model to generalize based on categorical datasets representingspatial variability. The framework is developed on a rich benchmark datasetfrom the US and evaluated on a smaller dataset collected by The NatureConservancy in Kenya. The LSTM model exhibits generalization performancethrough our TL technique. Results from this current experiment demonstrate theeffective predictive skill of forecasting streamflow responses when knowledgetransferring and static descriptors are used to improve hydrologic modelgeneralization in data-sparse regions.

Transfer learning to improve streamflow forecasts in data sparse regions