[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-125688-en":3,"doc-seo-125688-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},125688,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","Understanding molecular abundances in star-forming regions using interpretable machine learning","Astrochemical modelling of the interstellar medium often relies on complex computational codes with adjustable input parameters, but the precise mapping from inputs to molecular abundances remains difficult to interpret. This study applies SHapley Additive exPlanations (SHAP), an interpretable machine-learning method, to identify key physical parameters and their influence on abundance outputs. Neural-network emulators speed the evaluation of species abundances, enabling SHAP-based relationship analysis. Results show strong metallicity dependence for H2O and CO gas-phase abundances, distinct temperature regimes for NH3, and chemistry-linked correlations for HCN/HNC and HCN/CS ratios.","MNRAS 526, 404–422 (2023) [https://doi.org/10.1093/mnras/stad2814](https://doi.org/10.1093/mnras/stad2814)  \nAdvance Access publication 2023 September 14  \nUnderstanding molecular abundances in star-forming regions using interpretable machine learning  \nJohannes Heyl  , 1‹ Joshua Butterworth2 and Serena Viti1,2‹  \n1Department of Physics and Astronomy, University College London, Gower Street, WC1E 6BT London, UK  \n2Leiden Observatory, Leiden University, Huygens Laboratory, Niels Bohrweg 2, NL-2333 CA Leiden, the Netherlands  \nAccepted 2023 September 13. Received 2023 September 1; in original form 2023 May 14  \nABSTRACT  \nAstrochemical modelling of the interstellar medium typically makes use of complex computational codes with parameters whose values can be varied. It is not always clear what the exact nature of the relationship is between these input parameters and the output molecular abundances. In this work, a feature importance analysis is conducted using SHapley Additive exPlanations (SHAP), an interpretable machine learning technique, to identify the most important physical parameters as well as their relationship with each output. The outputs are the abundances of species and ratios of abundances. In order to reduce the time taken for this process, a neural network emulator is trained to model each species’ output abundance and this emulator is used to perform the interpretable machine learning. SHAP is then used to further explore the relationship between the physical features and the abundances for the various species and ratios we considered. H2 O and CO’s gas phase abundances are found to strongly depend on the metallicity. NH3 has a strong temperature dependence, with there being two temperature regimes (\u003C100 K and> 100 K) . By analysing the chemical network, we relate this to the chemical reactions in our network and find the increased temperature results in increased efficiency of destruction pathways. We investigate the HCN/HNC ratio and show that it can be used as a cosmic thermometer, agreeing with the literature. This ratio is also found to be correlated with the metallicity. The HCN/CS ratio serves as a density tracer, but also has three separate temperature-dependence regimes, which are linked to the chemistry of the two molecules.  \nKey words: astrochemistry–methods: statistical–stars: abundances.  \n1 INTRODUCTION  \nModelling the interstellar medium and star formation is often a complex matter. This is normally done using computational codes that take in a number of physical parameters and use these to integrate the system of coupled ordinary differential equations that represent a chemical network (Taquet, Ceccarelli & Kahane 2012; Ruaud, Wakelam & Hersant 2016; Holdship et al. 2017). However, duetothe non-linear nature of the chemistry, it is often unclear what the exact relationship is between the initial parameters and the output chemical abundances of the molecules of interest. This is often complicated by the fact that the various parameters have differing effects on the output abundances for different ranges.  \nIt has been customary in astrochemistry to consider grids of models in which the various parameters are varied (Taquet et al. 2012; Tunnard & Greve 2016; Viti 2017; Bianchi et al. 2019; James et al. 2020; Holdship & Viti 2022; Heyl et al. 2023) . The time-consuming and computationally expensive nature of many computational codes often limits the total number of model evaluations possible. This makes drawing conclusions about the importance of various parameters difficult. In this work, we look to address both of these issues. We  \n􀀂 E-mail: johannes.heyl.19@ucl.ac.uk (JH); viti@strw.leidenuniv.nl (SV)  \nmake use of SHapley Additive exPlanations (SHAP) (Lundberg & Lee 2017) to help improve our understanding of a chemical code. SHAP provides us with a means of understanding why a machine learning model outputs a particular value. By considering various combinations of inputs and outputs, these tech","cbCaicYP5AV3NvWv","https://ap.wps.com/l/cbCaicYP5AV3NvWv","pdf",4358702,1,19,"English","en",105,"# Abstract\n# 1 Introduction\n## Interpreting input–output relationships in chemical modelling\n## Using SHAP and statistical emulation to improve efficiency\n## Emulators for accelerating forward modelling","[{\"question\":\"Why is it difficult to connect input parameters to molecular abundance outputs in astrochemical modelling?\",\"answer\":\"Because chemical networks are strongly nonlinear, different parameters affect abundance outputs differently across parameter ranges, making the input–output relationship hard to determine.\"},{\"question\":\"How does this work make interpretable machine learning computationally feasible?\",\"answer\":\"It trains neural-network emulators to rapidly predict each species’ final abundance from the chemical-code inputs, then applies SHAP to these emulator outputs for interpretability.\"},{\"question\":\"What physical trends were found for key molecules and abundance ratios?\",\"answer\":\"H2O and CO gas-phase abundances strongly depend on metallicity. NH3 shows two temperature regimes, and HCN/HNC behaves as a cosmic thermometer with metallicity correlation, while HCN/CS acts as a density tracer with multiple temperature-dependent regimes linked to molecular chemistry.\"}]","Understanding molecular abundances in star-forming regions using interpretable machine learning | PDF",1785900676,48,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"understanding-molecular-abundances-in-star-forming-regions-using-interpretable-machine-learning-125688","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/understanding-molecular-abundances-in-star-forming-regions-using-interpretable-machine-learning-125688/125688/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why is it difficult to connect input parameters to molecular abundance outputs in astrochemical modelling?","Question",{"text":75,"@type":76},"Because chemical networks are strongly nonlinear, different parameters affect abundance outputs differently across parameter ranges, making the input–output relationship hard to determine.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does this work make interpretable machine learning computationally feasible?",{"text":80,"@type":76},"It trains neural-network emulators to rapidly predict each species’ final abundance from the chemical-code inputs, then applies SHAP to these emulator outputs for interpretability.",{"name":82,"@type":73,"acceptedAnswer":83},"What physical trends were found for key molecules and abundance ratios?",{"text":84,"@type":76},"H2O and CO gas-phase abundances strongly depend on metallicity. NH3 shows two temperature regimes, and HCN/HNC behaves as a cosmic thermometer with metallicity correlation, while HCN/CS acts as a density tracer with multiple temperature-dependent regimes linked to molecular chemistry.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},"General","general"]