scholarly journals Modelling Claim Frequency in Insurance Using Count Models

2021 ◽  
pp. 14-20
Author(s):  
A. Adetunji Ademola ◽  
Shamsul Rijal Muhammad Sabri
Keyword(s):  
Bayesian Information Criterion ◽  
Regression Models ◽  
Negative Binomial ◽  
Predictive Ability ◽  
Information Criterion ◽  
Information Criteria ◽  
Binomial Model ◽  
Count Models ◽  
Claim Frequency ◽  
Insurance Data

Background: In modelling claim frequency in actuary science, a major challenge is the number of zero claims associated with datasets. Aim: This study compares six count regression models on motorcycle insurance data. Methodology: The Akaike Information Criteria (AIC) and the Bayesian Information Criterion (BIC) were used for selecting best models. Results: Result of analysis showed that the Zero-Inflated Poisson (ZIP) with no regressors for the zero component gives the best predictive ability for the data with the least BIC while the classical Negative Binomial model gives the best result for explanatory purpose with the least AIC.

scholarly journals A Study of Count Regression Models for Mortality Rate

CAUCHY ◽  
2021 ◽  
Vol 7 (1) ◽  
pp. 142-151
Author(s):  
Anwar Fitrianto
Keyword(s):  
Mortality Rate ◽  
Regression Model ◽  
Count Data ◽  
Bayesian Information Criterion ◽  
Regression Models ◽  
Negative Binomial ◽  
Negative Binomial Regression ◽  
Information Criterion ◽  
Poisson Regression Model ◽  
Binomial Regression

This paper discusses how overdispersed count data to be fit. Poisson regression model, Negative Binomial 1 regression model (NEGBIN 1) and Negative Binomial regression 2 (NEGBIN 2) model were proposed to fit mortality rate data. The method used is comparing the values of Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) to find out which method suits the data the most. The results show that the data indeed display higher variability. Among the three models, the model preferred is NEGBIN 1 model.

scholarly journals Model Selection Procedures in Bounds Test of Cointegration: Theoretical Comparison and Empirical Evidence

Economies ◽  
2020 ◽  
Vol 8 (2) ◽  
pp. 49 ◽  
Author(s):  
Waqar Badshah ◽  
Mehmet Bulut
Keyword(s):  
Model Selection ◽  
Akaike Information Criterion ◽  
Bayesian Information Criterion ◽  
Selection Process ◽  
Information Criterion ◽  
Small Sample ◽  
Information Criteria ◽  
Path Model ◽  
Sample Sizes ◽  
Bounds Test

Only unstructured single-path model selection techniques, i.e., Information Criteria, are used by Bounds test of cointegration for model selection. The aim of this paper was twofold; one was to evaluate the performance of these five routinely used information criteria {Akaike Information Criterion (AIC), Akaike Information Criterion Corrected (AICC), Schwarz/Bayesian Information Criterion (SIC/BIC), Schwarz/Bayesian Information Criterion Corrected (SICC/BICC), and Hannan and Quinn Information Criterion (HQC)} and three structured approaches (Forward Selection, Backward Elimination, and Stepwise) by assessing their size and power properties at different sample sizes based on Monte Carlo simulations, and second was the assessment of the same based on real economic data. The second aim was achieved by the evaluation of the long-run relationship between three pairs of macroeconomic variables, i.e., Energy Consumption and GDP, Oil Price and GDP, and Broad Money and GDP for BRICS (Brazil, Russia, India, China and South Africa) countries using Bounds cointegration test. It was found that information criteria and structured procedures have the same powers for a sample size of 50 or greater. However, BICC and Stepwise are better at small sample sizes. In the light of simulation and real data results, a modified Bounds test with Stepwise model selection procedure may be used as it is strongly theoretically supported and avoids noise in the model selection process.

scholarly journals Using Model Selection Criteria to Choose the Number of Principal Components

2021 ◽  
Vol 20 (3) ◽  
pp. 450-461
Author(s):  
Stanley L. Sclove
Keyword(s):  
Model Selection ◽  
Principal Components ◽  
Bayesian Information Criterion ◽  
Selection Criteria ◽  
Information Criterion ◽  
Information Criteria ◽  
Akaike's Information Criterion ◽  
Model Selection Criteria ◽  
Adequate Number ◽  
Number Of Principal Components

AbstractThe use of information criteria, especially AIC (Akaike’s information criterion) and BIC (Bayesian information criterion), for choosing an adequate number of principal components is illustrated.

scholarly journals Sensitivity and specificity of information criteria

2019 ◽  
Vol 21 (2) ◽  
pp. 553-565 ◽  
Author(s):  
John J Dziak ◽  
Donna L Coffman ◽  
Stephanie T Lanza ◽  
Runze Li ◽  
Lars S Jermiin
Keyword(s):  
Bayesian Information Criterion ◽  
Information Criterion ◽  
Information Criteria ◽  
The Other ◽  
Biological Research ◽  
Ratio Test ◽  
Relative Importance ◽  
Fields Of Study ◽  
Alternative Perspective ◽  
Practical Implications

Abstract Information criteria (ICs) based on penalized likelihood, such as Akaike’s information criterion (AIC), the Bayesian information criterion (BIC) and sample-size-adjusted versions of them, are widely used for model selection in health and biological research. However, different criteria sometimes support different models, leading to discussions about which is the most trustworthy. Some researchers and fields of study habitually use one or the other, often without a clearly stated justification. They may not realize that the criteria may disagree. Others try to compare models using multiple criteria but encounter ambiguity when different criteria lead to substantively different answers, leading to questions about which criterion is best. In this paper we present an alternative perspective on these criteria that can help in interpreting their practical implications. Specifically, in some cases the comparison of two models using ICs can be viewed as equivalent to a likelihood ratio test, with the different criteria representing different alpha levels and BIC being a more conservative test than AIC. This perspective may lead to insights about how to interpret the ICs in more complex situations. For example, AIC or BIC could be preferable, depending on the relative importance one assigns to sensitivity versus specificity. Understanding the differences and similarities among the ICs can make it easier to compare their results and to use them to make informed decisions.

scholarly journals Sensitivity and Specificity of Information Criteria

2018 ◽  
Author(s):  
John J Dziak ◽  
Donna L Coffman ◽  
Stephanie T Lanza ◽  
Runze Li ◽  
Lars Sommer Jermiin
Keyword(s):  
Bayesian Information Criterion ◽  
Information Criterion ◽  
Information Criteria ◽  
The Other ◽  
Biological Research ◽  
Ratio Test ◽  
Relative Importance ◽  
Fields Of Study ◽  
Alternative Perspective ◽  
Practical Implications

Information criteria (ICs) based on penalized likelihood, such as Akaike's Information Criterion (AIC), the Bayesian Information Criterion (BIC), and sample-size-adjusted versions of them, are widely used for model selection in health and biological research. However, different criteria sometimes support different models, leading to discussions about which is the most trustworthy. Some researchers and fields of study habitually use one or the other, often without a clearly stated justification. They may not realize that the criteria may disagree. Others try to compare models using multiple criteria but encounter ambiguity when different criteria lead to substantively different answers, leading to questions about which criterion is best. In this paper we present an alternative perspective on these criteria that can help in interpreting their practical implications. Specifically, in some cases the comparison of two models using ICs can be viewed as equivalent to a likelihood ratio test, with the different criteria representing different alpha levels and BIC being a more conservative test than AIC. This perspective may lead to insights about how to interpret the ICs in more complex situations. For example, AIC or BIC could be preferable, depending on the relative importance one assigns to sensitivity versus specificity. Understanding the differences and similarities among the ICs can make it easier to compare their results and to use them to make informed decisions.

scholarly journals Identifying and Predicting Major Factors Affecting the Suicides in Sri Lanka

2019 ◽  
pp. 1-7
Author(s):  
S. M. M. Lakmali ◽  
L. S. Nawarathna
Keyword(s):  
Sri Lanka ◽  
Prediction Accuracy ◽  
Regression Models ◽  
Negative Binomial ◽  
Information Criterion ◽  
Data Set ◽  
Factors Affecting ◽  
Married People ◽  
Civil Status ◽  
Major Factors

Aims: Identifying factors related to suicide and the prediction of future suicides are very important because suicide becomes a significant factor that engaged with education, social status, age, gender and many other factors. Therefore, the main objective of this study is to find the civil and education factors effecting on suicidal attempts in Sri Lanka and propose a model to predict the future suicides. Study Design: Statistical analysis with descriptive analysis and proposing models for predicting future suicides. Place and Duration of Study: Data collected from the Department of Police, Sri Lanka, between January 2006 and December 2016. Methodology: Data set has separated into two categories namely ‘civil data’ and ‘educational data’. We modeled the data from 2006 to 2011 and the data from 2014 to 2016 were used for model validation purposes. Quasi Poisson and negative binomial regression models were fitted to identify the major factors affecting suicide in both categories. Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) values were used to select the best model. Further, the Mean Absolute Percentage Deviation (MAPD) and Symmetric Mean Absolute Percent Error (SMAPE) were calculated to find the prediction accuracy of the proposed models. Results: For both regression models, the variables age, gender and level of education are significant for the models fitted for educational data, and civil status and gender are significant in the civil status dataset. According to the analysis, highest suicides were recorded for the age groups 21-30 and over 61 males, minimally-educated and married people. By considering the MAPD values, the prediction accuracy of both Quasi Poisson models and Negative binomial models were above 99%. But the negative binomial model is the best model because of the comparable high accuracy than the other model. A considerable reduction in suicides was obtained in 2010, due to the peaceful situation in Sri Lanka after the civil war. It is observed that by paying special attention to teenagers, old-aged and married people can reduce the number of suicides.

scholarly journals Sensitivity and specificity of information criteria

2015 ◽  
Author(s):  
John J. Dziak ◽  
Donna L. Coffman ◽  
Stephanie T. Lanza ◽  
Runze Li
Keyword(s):  
Model Selection ◽  
Sample Size ◽  
Likelihood Ratio Test ◽  
Bayesian Information Criterion ◽  
Information Criterion ◽  
Information Criteria ◽  
Ratio Test ◽  
Relative Importance ◽  
Missed Opportunities ◽  
High Bias

Choosing a model with too few parameters can involve making unrealistically simple assumptions and lead to high bias, poor prediction, and missed opportunities for insight. Such models are not flexible enough to describe the sample or the population well. A model with too many parameters can t the observed data very well, but be too closely tailored to it. Such models may generalize poorly. Penalized-likelihood information criteria, such as Akaike's Information Criterion (AIC), the Bayesian Information Criterion (BIC), the Consistent AIC, and the Adjusted BIC, are widely used for model selection. However, different criteria sometimes support different models, leading to uncertainty about which criterion is the most trustworthy. In some simple cases the comparison of two models using information criteria can be viewed as equivalent to a likelihood ratio test, with the different models representing different alpha levels (i.e., different emphases on sensitivity or specificity; Lin & Dayton 1997). This perspective may lead to insights about how to interpret the criteria in less simple situations. For example, AIC or BIC could be preferable, depending on sample size and on the relative importance one assigns to sensitivity versus specificity. Understanding the differences among the criteria may make it easier to compare their results and to use them to make informed decisions.

scholarly journals Modelling the Number of Household Visit to Health Care Centres in Some Nigeria Communities Using Count Data Regression Models

2021 ◽  
Author(s):  
Samuel Olorunfemi Adams ◽  
Davies Abiodun Obaromi ◽  
Rauf Ibrahim Rauf
Keyword(s):  
Health Care ◽  
Community Health ◽  
Regression Models ◽  
Negative Binomial ◽  
Information Criterion ◽  
Community Health Care ◽  
Community Health Centres ◽  
Health Centres ◽  
Household Members ◽  
Household Visit

Introduction: The need to model the impact of some demographic indicators on the frequency of household visits to healthcare centres in Nigeria's community is very important for preventing and spreading community diseases. This study aimed to investigate the effect of the patents' age, gender, marital status, type of illness and amount spent on the frequency of visits to community health care centres in Nigeria and to compared Negative Binomial Regression (NBR) and Generalized Poisson Regression (GPR) models to determine the preferred count regression model for the number of household visits to health centres in some communities in Nigeria. Methods: Survey of 132640 households in some Nigeria communities obtained from the 2018/2019 Nigeria Living Standard Survey (NLSS) were extracted from the National Bureau of Statistics (NBS) in collaboration with the World Bank. The Negative Binomial and Generalised Poisson regression models were used to investigate the five demographic variables on the frequency of visit to the community health centres. The performance of the count regression model was assessed using the Chi-square -2log Likelihood Statistic (2logL), Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) selection criteria. Results: Findings from the study showed that the type of illness and amount spent has a significantly positive effect on the number of household members' visits to the community health care centres in Nigeria while age, gender, and marital status was discovered to have a negative effect on the number of household members' visits to the community health care centres in Nigeria. Conclusion: The Nigeria Government, health centre management and community healthcare service providers' need to be aware that the amount spent and the nature of illness determines the level of health care services utilisation in the Nigeria community, hence the need for the drastic reduction in charges to encourage a household visit to the community health centres whenever the need arises.

scholarly journals Modelling Claim Frequency in Vehicle Insurance

2016 ◽  
Vol 64 (2) ◽  
pp. 683-689
Author(s):  
Jiří Valecký
Keyword(s):  
Linear Models ◽  
Negative Binomial ◽  
Poisson Model ◽  
Binomial Model ◽  
Insurance Company ◽  
Linear Effect ◽  
New Findings ◽  
Insurance Portfolio ◽  
Systematic Component ◽  
Claim Frequency

The paper is focused on modelling claim frequency and extends the work of Kafková and Křivánková, 2014 (Kafková, S., Křivánková, L. 2014. Generalized linear models in vehicle insurance. Acta universitatis agriculturae et silviculturae mendelianae brunensis, 62(2): 383–388). We showed that overdispersion, non-linear systematic component and interacted rating factors should be considered when the claim frequency is modelled. We detected overdispersion in the Poisson model and employed the negative-binomial model to show that considering heterogeneity over insurance policies yields better fit of the model. We also analysed the linear effect of continuous rating factors and their mutual influences. We showed that non-linearity and interactions between rating factors yield the better fit of the model, as well as new findings related to the analysis of claim frequency. All empirical models were estimated on the insurance portfolio of Czech insurance company collected during the years 2004–2008.

Sign in / Sign up

Export Citation Format

Share Document