XGBoost is recognized as an algorithm with exceptional predictive capacity. Models for a binary response indicating the existence of accident claims versus no claims can be used to identify the determinants of traffic accidents. This study compared the relative performances of logistic regression and XGBoost approaches for predicting the existence of accident claims using telematics data. The dataset contained information from an insurance company about the individuals' driving patterns-including total annual distance driven and percentage of total distance driven in urban areas. Our findings showed that logistic regression is a suitable model given its interpretability and good predictive capacity. XGBoost requires numerous model-tuning procedures to match the predictive performance of the logistic regression model and greater effort as regards to interpretation.Risks 2019, 7, 70 2 of 16 among many others. Here, pay-as-you-drive (PAYD) insurance schemes represent an alternative method for pricing premiums based on personal mileage travelled and driving behaviors. Guillen et al. (2019), Verbelen et al. (2018, and Pérez-Marín and Guillén (2019) showed the potential benefits of analyzing telematics information when calculating motor insurance premiums. Gao and Wüthrich (2019) analyzed high-frequency GPS location data (second per second) of individual car drivers and trips. Gao and Wüthrich (2018) and investigated the predictive power of covariates extracted from telematics car driving data using the speed-acceleration heatmaps proposed by Wüthrich (2017). Further, Hultkrantz et al. (2012) highlighted the importance of PAYD insurance plans insofar as they allow insurance companies to personalize premium calculation and, so, charge fairer rates.The rest of this paper is organized as follows. First, the notation is introduced and the logistic regression and XGBoost methods are outlined. Second, our dataset is described and some descriptive statistics are provided. Third, the results of our comparisons in both a training and a testing sample are reported. Finally, following the conclusion, some practical suggestions are offered about the feasibility of applying new machine learning methods to the field of insurance.

In order to help in designing an accurate pension reform, we determine the resource allocation in an endogenous fertility model that generates an endogenous demographic transition by means of distinguishing between female and male labor. We analyze the problem of the optimal solution and characterize the decentralization of the first best. We show that a pension policy linking pension benefits to the number of children acts as a corrective tax system able to restore both the optimal capital stock and the optimal rate of population growth as a single instrument. We also show that neither a Beveridgean pension scheme nor a Bismarckian one can decentralize the first best.

Evidence on trends in prevalence of disease and disability can clarify whether countries are experiencing a compression or expansion of morbidity. An expansion of morbidity, as indicated by disease, has appeared in Europe and other developed regions. It is likely that better treatment, preventive measures, and increases in education levels have contributed to the declines in mortality and increments in life expectancy. This paper examines whether there has been an expansion of morbidity in Catalonia (Spain). It uses trends in mortality and morbidity and links these with survival to provide estimates of life expectancy with and without diseases and mobility limitations. We use a repeated cross-sectional health survey carried out in 1994 and 2011 for measures of morbidity, and information from the Spanish National Statistics Institute for mortality. Our findings show that at age 65 the percentage of life with disease increased from 52 to 70 % for men, and from 56 to 72 % for women; the expectation of life with mobility limitations increased from 24 to 30 % for men and from 40 to 47 % for women between 1994 and 2011. These changes were attributable to increases in the prevalence of diseases and moderate mobility limitation. Overall, we find an expansion of morbidity along the period. Increasing survival among people with diseases can lead to a higher prevalence of diseases in the older population. Higher prevalence of health problems can lead to greater pressure on the health care system and a growing burden of disease for individuals.

OBJECTIVE To analyze the prevalence of individuals at risk of dependence and its associated factors.METHODS The study was based on data from the Catalan Health Survey, Spain conducted in 2010 and 2011. Logistic regression models from a random sample of 3,842 individuals aged ≥ 15 years were used to classify individuals according to the state of their personal autonomy. Predictive models were proposed to identify indicators that helped distinguish dependent individuals from those at risk of dependence. Variables on health status, social support, and lifestyles were considered.RESULTS We found that 18.6% of the population presented a risk of dependence, especially after age 65. Compared with this group, individuals who reported dependence (11.0%) had difficulties performing activities of daily living and had to receive support to perform them. Habits such as smoking, excessive alcohol consumption, and being sedentary were associated with a higher probability of dependence, particularly for women.CONCLUSIONS Difficulties in carrying out activities of daily living precede the onset of dependence. Preserving personal autonomy and function without receiving support appear to be a preventive factor. Adopting an active and healthy lifestyle helps reduce the risk of dependence.

BackgroundHealth expectancies vary worldwide according to socioeconomic status (SES), with health disadvantages being evident among lower SES groups. Using educational attainment as a proxy of SES, we seek to identify trends in SES differentials in health by gender, with a particular focus on individuals with low educational attainment in the adult Catalan population (Spain) aged 55 or older.MethodsUsing cross-sectional data for 1994 and 2010-2014 drawn from the Catalan Health Survey, we examined three health indicators to document social health inequalities: self-perceived health, functional limitations, and restrictions on activities of daily living (ADL). We applied logistic models for each indicator, controlling for sociodemographic characteristics, health coverage and health behaviours.ResultsAmong the less-educated, females presented a greater improvement in their self-perceived health over time than did their male counterparts, there being no significant variations among the medium/high educated. Regardless of education, males showed an increase in the prevalence of functional problems (as did the women, but the increase was not statistically significant). Both genders presented a higher prevalence of limitations when performing ADL in the second time period. The gender health gap was reduced slightly both for the low and the medium/high educated, expect in the case of ADL restrictions. Health and functioning differences by education level persisted, but showed significant signs of reduction.ConclusionsLess-educated females constitute the most disadvantaged group in terms of health and personal autonomy, though there are encouraging signs that the gap is closing both in terms of gender and level of education. Health policymakers need to devote particular attention to the aging population with low SES, especially to women. Public programmes promoting greater protection and equity, while fostering preventive and healthy practices, need to target the most underprivileged.

This study examines the influence of risk factors on mental well-being at older ages focusing on the level of rurality of the living environment. We used cross-sectional, nationally representative survey data for Catalonia (Spain) from 2015 to 2017 to explain the mental well-being of the population aged 65 years and over. Based on a sample of 2621 individuals, we created a score of current mental well-being using the Short Warwick-Edinburgh Mental Well-being Scale (SWEMWBS). Using logistic regression and non-parametric tests, we identified the sociodemographic, health and lifestyle variables which, in combination with the features of the rural and urban settings of the living space, were associated with poor SWEMWBS scores. Our results reveal that adequate social support is linked to expectations of good mental well-being in later life. Poor self-perceived health and ageing limitations are associated with less deterioration of the well-being for the elderly living in rural areas, whereas living in urban areas is linked to a higher risk of suffering from emotional distress attributable to economic difficulties or low educational attainment. Incentivizing older people to live in rural environments could result in greater well-being in the last stages of life; appropriate prospective studies are needed to test this positive outcome.

Purpose: This paper aims at analyzing the skills and competences acquired by students during their university studies, in order to identify the extent to which they fit into society's labor market demands. Research limitations/implications:The database used refers only to one Faculty of the University of Barcelona. It would be interesting to collect more data and extend the results to a more general context. Practical implications:It might be interesting both for firms and universities to move closer to one another in a cooperative way. More programs are needed to further -612-Intangible Capital -http://dx.doi. org/10.3926/ic.625 enhance the employability of graduates and encourage the mobility of teachers in order for them to spend some time performing their research directly at a firm.Originality/value: This paper explores the immediate and direct relationship established between the university and the labor market through the training period students make during the final stages of their studies.

From 1850 to 2000, in Western European countries life expectancy rose from 30-40 to 80 years and the average number of children per woman fell from 4 to 5 children to slightly more than one. To gauge the economic consequences of these demographic trends, we implement an overlapping generations model with heterogeneity by level of education in which individuals optimally decide their consumption of market-and home-produced goods as well as the time spent on paid and unpaid work. We find that around 17% of the observed increase in per-capita income growth from 1850 to 2000 was due to the demographic transition. Around 50% of the demo-This project has received institutional support from the European Union's Seventh Framework Programme for research, technological development and demonstration (AGENTA project, grant agreement No: 613247). SERIEs (2018) 9:27-64 graphic contribution is explained by the increase in the average productivity per worker (productivity component), which arises from the change in the population's age structure and the rise in households' saving rate. The remaining 50% is explained by the higher growth rate of workers relative to the total population (translation component). Electronic supplementary material The online version of this article (