Claim frequency models in vehicle insurance based on GLM

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27510%2F18%3A10240119" target="_blank" >RIV/61989100:27510/18:10240119 - isvavai.cz</a>

Výsledek na webu

<a href="https://mme2018.fm.vse.cz/wp-content/uploads/2018/09/MME2018-Electronic_proceedings.pdf" target="_blank" >https://mme2018.fm.vse.cz/wp-content/uploads/2018/09/MME2018-Electronic_proceedings.pdf</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Claim frequency models in vehicle insurance based on GLM

Popis výsledku v původním jazyce

Within non-life insurance pricing determinated by insurance premium precedes and essential part represented models of claim frequency and claim severity. Theese models are usually modelling by generalized linear models. This paper is focused on estimation of claim frequency and extends the work ŠPAČKOVÁ, Adéla. Estimation of claim frequency by generalized linear models, 2017, s. 821 - 830. ISBN 978-80-248-4138-0. Regression analysis allows the identification of the risk factors and the prediction of the expected frequency of claims given the characteristics of policyholders. It depends on many individual rating factors (e.g. based on individual characteristics of vehicle and driver). The aim of this paper is to find out ideally suited model for estimation claim frequency based on these risk factors. All empirical models are estimated on the real world sample data of czech insurance company collected during the years 2005 - 2010. Parameters of model are estimated by maximum likelihood method at standard level of significant 0,05. Verification of the model parameters is perfomed by a Wald test. Comparison models with different predictor variables is established by analysis of deviance residuals, Akaike information criterion (AIC) and Bayesian information criterion (BIC). Based on these comparison the ideally suited model is choosen. All calculations are computed in statistical software STATA 14.00.

Název v anglickém jazyce

Claim frequency models in vehicle insurance based on GLM

Popis výsledku anglicky

Within non-life insurance pricing determinated by insurance premium precedes and essential part represented models of claim frequency and claim severity. Theese models are usually modelling by generalized linear models. This paper is focused on estimation of claim frequency and extends the work ŠPAČKOVÁ, Adéla. Estimation of claim frequency by generalized linear models, 2017, s. 821 - 830. ISBN 978-80-248-4138-0. Regression analysis allows the identification of the risk factors and the prediction of the expected frequency of claims given the characteristics of policyholders. It depends on many individual rating factors (e.g. based on individual characteristics of vehicle and driver). The aim of this paper is to find out ideally suited model for estimation claim frequency based on these risk factors. All empirical models are estimated on the real world sample data of czech insurance company collected during the years 2005 - 2010. Parameters of model are estimated by maximum likelihood method at standard level of significant 0,05. Verification of the model parameters is perfomed by a Wald test. Comparison models with different predictor variables is established by analysis of deviance residuals, Akaike information criterion (AIC) and Bayesian information criterion (BIC). Based on these comparison the ideally suited model is choosen. All calculations are computed in statistical software STATA 14.00.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

50206 - Finance

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2018

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Mathematical Methods in Economics : MME 2018 : 36th international conference : September 12-14, 2018, Jindřichův Hradec

ISBN

978-80-7378-371-6

ISSN

—

e-ISSN

—

Počet stran výsledku

6

Strana od-do

555-560

Název nakladatele

MATFYZPRESS

Místo vydání

Praha

Místo konání akce

Jindřichův Hradec

Datum konání akce

12. 9. 2018

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000507455300096