[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-124888-en":3,"doc-seo-124888-105":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},124888,137441390410,"Hazel","https://ap-avatar.wpscdn.com/avatar/2000252f4ab5702993?_k=1776741390130283984",8,"Research & Report","Business Sector Classification and Beyond Using Machine Learning - Statistical Note 18 - 2024","This statistical note presents Banco de España’s Central Balance Sheet Data Office work on sectorisation and classification of holding companies using supervised machine learning. It describes an automated procedure to distinguish holding companies and head offices, supported by financial-statement ratios and linked to CNAE codes 6420 and 7010. It then builds models to classify entities into financial versus non-financial institutional sectors for National Accounts. Quality control includes automatic classification of 8,500 firms and manual review of around 5,300 when CNAE coding differs.","BUSINESS SECTOR CLASSIFICATION AND BEYOND USING MACHINE LEARNING  \n2024  \nNotas Estadísticas N.º 18  \nAlejandro Morales Fernández  \nCONTENTS  \nAbstract 4  \nResumen 5  \n1 Introduction 7  \n1.1 Initial motivations 7  \n1.2 Previous work carried out by other central banks 8  \n1.3 Preprocessing and variable selection 9  \n1.3.1 Data Engineering 9  \n1.3.2 Feature Engineering 10  \n2 Construction of the supervised business sector classification model 12  \n2.1 Business rules associated with holding companies and head offices 12  \n2.2 Final distinction between holding companies and head offices based on the employment business rule 13  \n2.3 Final results for the business sector model 13  \n2.4 Final variables and model interpretation of the business sector model with Shapley Values 16  \n2.5 Review tasks performed by business staff 17  \n2.5.1 First Review (CBA) 18  \n2.5.2 Second Review (CBB) 18  \n3 Construction of the supervised institutional sector classification model 19  \n3.1 Business rules associated with financial holding companies and financial head offices 19  \n3.2 Challenges and focus 20  \n3.3 Data Engineering 20  \n3.4 Feature Engineering 22  \n3.5 Final model 22  \n3.6 Variable interpretation 22  \n3.7 Review tasks performed by business staff 26  \n4 Conclusions and lessons learnt 26  \n4.1 Conclusions 26  \n4.2 Lessons learnt 27  \n4.3 Next steps 27  \n5 Annex: technical details of the models 27  \n5.1 Variable selection and feature engineering 27  \n5.1.1 Elimination of variables due to high correlations 28  \n5.1.2 Categorical variable treatment 28  \n5.1.3 Missing values treatment 28  \n5.1.4 Variable selection and importance ranking using Random Forest and SHAP values 29  \n5.1.5 Selection of the number of variables 29  \n5.2 Preliminary steps carried out prior to model construction 29  \n5.2.1 Data partitioning and first models with training-test split and cross-validation 29  \n5.2.2 Decision Trees and Random Forests 30  \n5.2.3 Application of other classification models 30  \n5.2.4 Sample balancing 31  \n5.3 Retraining the business sector model with corrected training data 31  \n5.4 Parameter grid 32  \n5.5 Business rules taught to the algorithm 32  \n5.6 Interpretation and impact of variables in the model 33  \nReferences 34  \nTechnical Glossary 35  \nStatistical notes published 37  \nBUSINESS SECTOR CLASSIFICATION AND BEYOND USING MACHINE LEARNING  \nABSTRACT  \nThis statistical note presents the work carried out last year by the Banco de España’s Central Balance Sheet Data Office (CBSO) on the sectorisation and classification of holding companies using machine learning. This work has also been presented, in July 2023, at the World Statistics Congress (WSC) in Ottawa, organised by the International Statistics Institute (ISI), and this note is part of a series of talks on central banks organised by the Irving Fisher Committee (IFC) at the same congress.  \nThe work presented can be divided into two parts: first, obtaining an automated procedure to help distinguish companies that, given their economic activity, are either holding companies or head offices. In other words, the aim of this work is to detect companies whose activities may come under codes 6420 or 7010 of the CNAE (Spanish National Classification of Economic Activities, equivalent to NACE, the statistical classification of economic activities in the European Community), by checking whether their data (mainly economic and financial ratios from their annual financial statements) suggest that they are or may be holding companies or head offices (whether or not they report such activities) . The second part of the work is the classification of holding companies and head offices into the financial or non-financial sectors (as required by the National Accounts), using the model and information generated by the first part of the project as a starting point.  \nArtificial intelligence – in particular supervised machine learning classification models – is used to perform both of these tasks. A super","cbCaif8kY4kt0Rdb","https://ap.wps.com/l/cbCaif8kY4kt0Rdb","pdf",1284628,1,38,"English","en",105,"# Introduction\n## Initial motivations\n## Previous work carried out by other central banks\n## Preprocessing and variable selection\n# Construction of the supervised business sector classification model\n## Business rules for holding companies and head offices\n## Final results and model interpretation with Shapley values\n## Review tasks performed by business staff\n# Construction of the supervised institutional sector classification model\n## Business rules for financial holdings and financial head offices\n## Challenges and focus\n## Data engineering and feature engineering\n## Model review tasks performed by business staff\n# Conclusions and lessons learnt\n## Conclusions\n## Lessons learnt\n## Next steps\n# Annex: technical details of the models\n## Variable selection and feature engineering","[{\"question\":\"How does the note distinguish holding companies from head offices?\",\"answer\":\"It uses an automated, supervised machine-learning procedure that checks whether companies’ financial-statement ratios indicate holding-company or head-office characteristics tied to CNAE codes 6420 and 7010.\"},{\"question\":\"What is the second modeling task described?\",\"answer\":\"The note constructs a supervised institutional sector classification model that assigns holding companies and head offices to financial or non-financial sectors, as required by the National Accounts.\"},{\"question\":\"How is model quality controlled when CNAE codes differ from original records?\",\"answer\":\"The proposed procedure combines automatic classification for firms where results match business rules (over 8,500) with manual review for the remaining cases (about 5,300).\"}]","Business Sector Classification and Beyond Using Machine Learning - Statistical Note 18 - 2024 | PDF",1785895236,96,{"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},"business-sector-classification-and-beyond-using-machine-learning-statistical-note-18-2024","",{"@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/business-sector-classification-and-beyond-using-machine-learning-statistical-note-18-2024/124888/",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},"How does the note distinguish holding companies from head offices?","Question",{"text":75,"@type":76},"It uses an automated, supervised machine-learning procedure that checks whether companies’ financial-statement ratios indicate holding-company or head-office characteristics tied to CNAE codes 6420 and 7010.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What is the second modeling task described?",{"text":80,"@type":76},"The note constructs a supervised institutional sector classification model that assigns holding companies and head offices to financial or non-financial sectors, as required by the National Accounts.",{"name":82,"@type":73,"acceptedAnswer":83},"How is model quality controlled when CNAE codes differ from original records?",{"text":84,"@type":76},"The proposed procedure combines automatic classification for firms where results match business rules (over 8,500) with manual review for the remaining cases (about 5,300).","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":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]