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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

In many astrophysical applications, the cost of solving a chemical networkrepresented by a system of ordinary differential equations (ODEs) growssignificantly with the size of the network, and can often represent asignificant computational bottleneck, particularly in coupled chemo-dynamicalmodels. Although standard numerical techniques and complex solutions tailoredto thermochemistry can somewhat reduce the cost, more recently, machinelearning algorithms have begun to attack this challenge via data-drivendimensional reduction techniques. In this work, we present a new class ofmethods that take advantage of machine learning techniques to reduce complexdata sets (autoencoders), the optimization of multi-parameter systems (standardbackpropagation), and the robustness of well-established ODE solvers to toexplicitly incorporate time-dependence. This new method allows us to find acompressed and simplified version of a large chemical network in asemi-automated fashion that can be solved with a standard ODE solver, whilealso enabling interpretability of the compressed, latent network. As a proof ofconcept, we tested the method on an astrophysically-relevant chemical networkwith 29 species and 224 reactions, obtaining a reduced but representativenetwork with only 5 species and 12 reactions, and a x65 speed-up.

Reducing the complexity of chemical networks via interpretable autoencoders