From point to probabilistic gradient boosting for claim frequency and severity prediction

<?xml version="1.0" encoding="UTF-8"?><modsCollection xmlns="http://www.loc.gov/mods/v3" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-8.xsd">
<mods version="3.8">
<titleInfo>
<title>From point to probabilistic gradient boosting for claim frequency and severity prediction</title>
</titleInfo>
<name type="personal" xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="MAPA20260002415">
<namePart>Côté, Marie-Pier</namePart>
<nameIdentifier>MAPA20260002415</nameIdentifier>
</name>
<name type="corporate" xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="MAPA20180008764">
<namePart>Springer</namePart>
<nameIdentifier>MAPA20180008764</nameIdentifier>
</name>
<typeOfResource>text</typeOfResource>
<genre authority="marcgt">periodical</genre>
<originInfo>
<place>
<placeTerm type="code" authority="marccountry">che</placeTerm>
</place>
<dateIssued encoding="marc">2025</dateIssued>
<issuance>serial</issuance>
</originInfo>
<language>
<languageTerm type="code" authority="iso639-2b">eng</languageTerm>
</language>
<physicalDescription>
<form authority="marcform">print</form>
</physicalDescription>
<abstract displayLabel="Summary">This survey paper provides a comprehensive review of point and probabilistic gradient boosting algorithms applied to insurance claim frequency and severity modelling. The authors compare leading boosting methodssuch as GBM, XGBoost, LightGBM, CatBoost, EGBM, XGBoostLSS, cyc-GBM and NGBoostusing public insurance datasets of varying size and complexity. The study evaluates computational efficiency, predictive performance, model adequacy, and calibration. Results show that LightGBM and XGBoostLSS are the most computationally efficient, CatBoost performs well with high-cardinality categorical variables, and probabilistic boosting methods can improve model adequacy without sacrificing accuracy. The paper highlights that no single method is universally superior, and offers actuarial guidance for choosing boosting algorithms for pricing and risk modelling</abstract>
<note type="statement of responsibility">Dominik Chevalier and Marie-Pier Côté</note>
<subject xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="MAPA20080592011">
<topic>Modelos actuariales</topic>
</subject>
<subject xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="MAPA20230006092">
<topic>Frecuencia siniestral</topic>
</subject>
<subject xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="MAPA20170005476">
<topic>Machine learning</topic>
</subject>
<subject xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="MAPA20080611613">
<topic>Modelos probabílisticos</topic>
</subject>
<subject xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="MAPA20080564322">
<topic>Tarificación</topic>
</subject>
<subject xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="MAPA20080592059">
<topic>Modelos predictivos</topic>
</subject>
<classification authority="">6</classification>
<relatedItem type="host">
<titleInfo>
<title>European Actuarial Journal</title>
</titleInfo>
<originInfo>
<publisher>Cham, Switzerland  : Springer Nature Switzerland AG,  2021-2022</publisher>
</originInfo>
<identifier type="local">MAP20220007085</identifier>
<part>
<text>11/08/2025 Volume 15 - Number 2 - August  2025 </text>
</part>
</relatedItem>
<recordInfo>
<recordContentSource authority="marcorg">MAP</recordContentSource>
<recordCreationDate encoding="marc">260206</recordCreationDate>
<recordChangeDate encoding="iso8601">20260211184311.0</recordChangeDate>
<recordIdentifier source="MAP">MAP20260003023</recordIdentifier>
<languageOfCataloging>
<languageTerm type="code" authority="iso639-2b">spa</languageTerm>
</languageOfCataloging>
</recordInfo>
</mods>
</modsCollection>