We consider linear random coefficient regression models, where the regressors are allowed to have a finite support. First, we investigate identification, and show that the means and the variances and covariances of the random coefficients are identified from the first two conditional moments of the response given the covariates if the support of the covariates, excluding the intercept, contains a Cartesian product with at least three points in each coordinate. We also discuss identification of higher-order mixed moments, as well as partial identification in the presence of a binary regressor. Next, we show the variable selection consistency of the adaptive LASSO for the variances and covariances of the random coefficients in finite and moderately high dimensions. This implies that the estimated covariance matrix will actually be positive semidefinite and hence a valid covariance matrix, in contrast to the estimate arising from a simple least squares fit. We illustrate the proposed method in a simulation study.

In this paper, we construct interpretable zero-inflated neural network models for modeling hospital admission counts related to respiratory diseases among a health-insured population and their dependants in the United States. In particular, we exemplify our approach by considering the zero-inflated Poisson neural network (ZIPNN), and we follow the combined actuarial neural network (CANN) approach for developing zero-inflated combined actuarial neural network (ZIPCANN) models for modeling admission rates, which can accommodate the excess zero nature of admission counts data. Furthermore, we adopt the LocalGLMnet approach (Richman & Wüthrich (2023). Scandinavian Actuarial Journal, 2023(1), 71–95.) for interpreting the ZIPNN model results. This facilitates the analysis of the impact of a number of socio-demographic factors on the admission rates related to respiratory disease while benefiting from an improved predictive performance. The real-life utility of the methodologies developed as part of this work lies in the fact that they facilitate accurate rate setting, in addition to offering the potential to inform health interventions.

The distribution-free chain ladder of Mack justified the use of the chain ladder predictor and enabled Mack to derive an estimator of conditional mean squared error of prediction for the chain ladder predictor. Classical insurance loss models, that is of compound Poisson type, are not consistent with Mack’s distribution-free chain ladder. However, for a sequence of compound Poisson loss models indexed by exposure (e.g., number of contracts), we show that the chain ladder predictor and Mack’s estimator of conditional mean squared error of prediction can be derived by considering large exposure asymptotics. Hence, quantifying chain ladder prediction uncertainty can be done with Mack’s estimator without relying on the validity of the model assumptions of the distribution-free chain ladder.

This article provides a structured description of openly available news topics and forecasts for armed conflict at the national and grid cell level starting January 2010. The news topics, as well as the forecasts, are updated monthly at conflictforecast.org and provide coverage for more than 170 countries and about 65,000 grid cells of size 55 × 55 km worldwide. The forecasts rely on natural language processing (NLP) and machine learning techniques to leverage a large corpus of newspaper text for predicting sudden onsets of violence in peaceful countries. Our goals are a) to support conflict prevention efforts by making our risk forecasts available to practitioners and research teams worldwide, b) to facilitate additional research that can utilize risk forecasts for causal identification, and c) to provide an overview of the news landscape.

This paper explores citizens’ stances toward the use of artificial intelligence (AI) in public services in Norway. Utilizing a social contract perspective, the study analyzes the government–citizen relationship at macro, meso, and micro levels. A prototype of an AI-enabled public welfare service was designed and presented to 20 participants who were interviewed to investigate their stances on the described AI use. We found a generally positive attitude and identified three factors contributing to this: (a) the high level of trust in government (macro level); (b) the balanced value proposition between individual and collective needs (meso level); and (c) the reassurance provided by having humans in the loop and providing transparency into processes, data, and model’s logic (microlevel). The findings provide valuable insights into citizens’ stances for socially responsible AI in public services. These insights can inform policy and guide the design and implementation of AI systems in the public sector by foregrounding the government–citizen relationship.

A survey of Hong Kong residents finds that public support for government technology, as understood through the concept of smart cities, is associated with concept-awareness and official communications. The statistical analysis identifies moderating effects attributable to personal social media use and controls for personal ideological views about scope of government intervention and perceived political legitimacy of smart city policies. The study builds on a growing body of empirical scholarship about public support for government technology, while also addressing a practical trend in urban governance: the growing sophistication of technologies like artificial intelligence and their use in strengthening government capacities. The Hong Kong case exemplifies ambitious investments in technology by governments and, at the time of the survey, relatively high freedom of political expression. The study’s findings help refine theories about state-society relations in the rapidly evolving context of technology for public sector use.

We present a detailed analysis of random motions moving in higher spaces with a natural number of velocities. In the case of the so-called minimal random dynamics, under some broad assumptions, we give the joint distribution of the position of the motion (for both the inner part and the boundary of the support) and the number of displacements performed with each velocity. Explicit results for cyclic and complete motions are derived. We establish useful relationships between motions moving in different spaces, and we derive the form of the distribution of the movements in arbitrary dimension. Finally, we investigate further properties for stochastic motions governed by non-homogeneous Poisson processes.

We consider the problem of obtaining effective representations for the solutions of linear, vector-valued stochastic differential equations (SDEs) driven by non-Gaussian pure-jump Lévy processes, and we show how such representations lead to efficient simulation methods. The processes considered constitute a broad class of models that find application across the physical and biological sciences, mathematics, finance, and engineering. Motivated by important relevant problems in statistical inference, we derive new, generalised shot-noise simulation methods whenever a normal variance-mean (NVM) mixture representation exists for the driving Lévy process, including the generalised hyperbolic, normal-gamma, and normal tempered stable cases. Simple, explicit conditions are identified for the convergence of the residual of a truncated shot-noise representation to a Brownian motion in the case of the pure Lévy process, and to a Brownian-driven SDE in the case of the Lévy-driven SDE. These results provide Gaussian approximations to the small jumps of the process under the NVM representation. The resulting representations are of particular importance in state inference and parameter estimation for Lévy-driven SDE models, since the resulting conditionally Gaussian structures can be readily incorporated into latent variable inference methods such as Markov chain Monte Carlo, expectation-maximisation, and sequential Monte Carlo.

This paper considers the family of invariant measures of Markovian mean-field interacting particle systems on a countably infinite state space and studies its large deviation asymptotics. The Freidlin–Wentzell quasipotential is the usual candidate rate function for the sequence of invariant measures indexed by the number of particles. The paper provides two counterexamples where the quasipotential is not the rate function. The quasipotential arises from finite-horizon considerations. However, there are certain barriers that cannot be surmounted easily in any finite time horizon, but these barriers can be crossed in the stationary regime. Consequently, the quasipotential is infinite at some points where the rate function is finite. After highlighting this phenomenon, the paper studies some sufficient conditions on a class of interacting particle systems under which one can continue to assert that the Freidlin–Wentzell quasipotential is indeed the rate function.

We show that for arrival processes, the ‘harmonic new better than used in expectation’ (HNBUE) (or ‘harmonic new worse than used in expectation’, HNWUE) property is a sufficient condition for inequalities between the time and customer averages of the system if the state of the system between arrival epochs is stochastically decreasing and convex and the lack of anticipation assumption is satisfied. HNB(W)UE is a wider class than NB(W)UE, being the largest of all available classes of distributions with positive (negative) aging properties. Thus, this result represents an important step beyond existing result on inequalities between time and customer averages, which states that for arrival processes, the NB(W)UE property is a sufficient condition for inequalities.

Current World Health Organization (WHO) reports claim a decline in COVID-19 testing and reporting of new infections. To discuss the consequences of ignoring severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection, the endemic characteristics of the disease in 2023 with the ones estimated before using 2022 data sets are compared. The accumulated numbers of cases and deaths reported to the WHO by the 10 most infected countries and global figures were used to calculate the average daily numbers of cases DCC and deaths DDC per capita and case fatality rates (CFRs = DDC/DCC) for two periods in 2023. In some countries, the DDC values can be higher than the upper 2022 limit and exceed the seasonal influenza mortality. The increase in CFR in 2023 shows that SARS-CoV-2 infection is still dangerous. The numbers of COVID-19 cases and deaths per capita in 2022 and 2023 do not demonstrate downward trends with the increase in the percentages of fully vaccinated people and boosters. The reasons may be both rapid mutations of the coronavirus, which reduced the effectiveness of vaccines and led to a large number of re-infections, and inappropriate management.

In this paper, we propose two new optimal reinsurance models in which both premium budget constraints and the reinsurer’s risk limits are taken into account. To be precise, we assume that the reinsurance premium has an upper bound, and that the admissible ceded loss functions have a pre-specified upper limit. Moreover, we assume that the reinsurance premium principle is calculated by the expected value premium principle. Under the optimality criteria of minimizing the value at risk and conditional value at risk of the insurer’s total risk exposure, we derive the explicit optimal reinsurance treaties, which are layer reinsurance treaties. A new approach is developed to construct the optimal reinsurance treaties. Comparisons with existing studies are also made. Finally, we provide a numerical study based on real data and an example to illustrate the proposed models and results. Our work provides a novel generalization of several known achievements in the literature.

Tuberculosis (TB) remains a global leading cause of death, necessitating an investigation into its unequal distribution. Sun exposure, linked to vitamin D (VD) synthesis, has been proposed as a protective factor. This study aimed to analyse TB rates in Spain over time and space and explore their relationship with sunlight exposure. An ecological study examined the associations between rainfall, sunshine hours, and TB incidence in Spain. Data from the National Epidemiological Surveillance Network (RENAVE in Spanish) and the Spanish Meteorological Agency (AEMET in Spanish) from 2012 to 2020 were utilized. Correlation and spatial regression analyses were conducted. Between 2012 and 2020, 43,419 non-imported TB cases were reported. A geographic pattern (north–south) and distinct seasonality (spring peaks and autumn troughs) were observed. Sunshine hours and rainfall displayed a strong negative correlation. Spatial regression and seasonal models identified a negative correlation between TB incidence and sunshine hours, with a four-month lag. A clear spatiotemporal association between TB incidence and sunshine hours emerged in Spain from 2012 to 2020. VD levels likely mediate this relationship, being influenced by sunlight exposure and TB development. Further research is warranted to elucidate the causal pathway and inform public health strategies for improved TB control.

From 2020 to December 2022, China implemented strict measures to contain the spread of severe acute respiratory syndrome coronavirus 2. However, despite these efforts, sustained outbreaks of the Omicron variants occurred in 2022. We extracted COVID-19 case numbers from May 2021 to October 2022 to identify outbreaks of the Delta and Omicron variants in all provinces of mainland China. We found that omicron outbreaks were more frequent (4.3 vs. 1.6 outbreaks per month) and longer-lasting (mean duration: 13 vs. 4 weeks per outbreak) than Delta outbreaks, resulting in a total of 865,100 cases, of which 85% were asymptomatic. Despite the average Government Response Index being 12% higher (95% confidence interval (CI): 9%, 15%) in Omicron outbreaks, the average daily effective reproduction number (Rt) was 0.45 higher (95% CI: 0.38, 0.52, p < 0.001) than in Delta outbreaks. Omicron outbreaks were suppressed in 32 days on average (95% CI: 26, 39), which was substantially longer than Delta outbreaks (14 days; 95% CI: 11, 19; p = 0.004). We concluded that control measures effective against Delta could not contain Omicron outbreaks in China. This highlights the need for continuous evaluation of new variants’ epidemiology to inform COVID-19 response decisions.

This paper considers the first passage times to constant boundaries and the two-sided exit problem for Lévy process with a characteristic exponent in which at least one of the two jumps having rational Laplace transforms. The joint distribution of the first passage times and undershoot/overshoot are obtained. The processes recover many models that have appeared in the literature such as the compound Poisson risk models, the perturbed compound Poisson risk models, and their dual ones. As applications, we obtain the solutions for popular path-dependent options such as lookback and barrier options in terms of Laplace transforms.

Hepatitis E virus (HEV) is a major cause of acute jaundice in South Asia. Gaps in our understanding of transmission are driven by non-specific symptoms and scarcity of diagnostics, impeding rational control strategies. In this context, serological data can provide important proxy measures of infection. We enrolled a population-representative serological cohort of 2,337 individuals in Sitakunda, Bangladesh. We estimated the annual risks of HEV infection and seroreversion both using serostatus changes between paired serum samples collected 9 months apart, and by fitting catalytic models to the age-stratified cross-sectional seroprevalence. At baseline, 15% (95 CI: 14–17%) of people were seropositive, with seroprevalence highest in the relatively urban south. During the study, 27 individuals seroreverted (annual seroreversion risk: 15%, 95 CI: 10–21%), and 38 seroconverted (annual infection risk: 3%, 95CI: 2–5%). Relying on cross-sectional seroprevalence data alone, and ignoring seroreversion, underestimated the annual infection risk five-fold (0.6%, 95 CrI: 0.5–0.6%). When we accounted for the observed seroreversion in a reversible catalytic model, infection risk was more consistent with measured seroincidence. Our results quantify HEV infection risk in Sitakunda and highlight the importance of accounting for seroreversion when estimating infection incidence from cross-sectional seroprevalence data.

In the third week of September 2022, an outbreak of measles was reported from a slum in Eastern Mumbai, India. We sought to investigate whether failure to vaccinate or vaccine failure was the cause. We constructed an epidemic curve, drew a spot map, and calculated the attack rate and case-fatality ratio. We calculated vaccine effectiveness (VE) for one and two doses of measles vaccine in an unmatched case–control study and did stratified analysis by sex, availability of vaccination card, and migrant status. We identified 358 cases and four deaths with a 11.3% attack rate and 1.1% case fatality, both being highest among 0–24-month-old boys. The epidemic curve suggested a propagated mode of spread. The VE for two doses was 64% (95% confidence interval (CI): 23–73%) among under-5-year-old children and 70% (95% CI: 28–88%) among 5–15-year-old children. Failure to vaccinate, consequent to the COVID-19 pandemic, and vaccine hesitancy might have led to the accumulation of susceptible children in the community. Additionally, the occurrence of case-patients among vaccinated suggests reduced VE, which needs further investigation into humoral and cell-mediated immunity as well as contributory factors including nutritional status. Outbreak response immunization to complete immunization of missed and dropout children was carried out to control the outbreak.

Most countries in Africa deployed digital solutions to monitor progress in rolling out COVID-19 vaccines. A rapid assessment of existing data systems for COVID-19 vaccines in the African region was conducted between May and July 2022, in 23 countries. Data were collected through interviews with key informants, identified among senior staff within Ministries of Health, using a semi-structured electronic questionnaire. At vaccination sites, individual data were collected in paper-based registers in five countries (21.7%), in an electronic registry in two countries (8.7%), and in the remaining 16 countries (69.6%) using a combination of paper-based and electronic registries. Of the 18 countries using client-based digital registries, 11 (61%) deployed the District Health Information System 2 Tracker, and seven (39%), a locally developed platform. The mean percentage of individual data transcribed in the electronic registries was 61% ± 36% standard deviation. Unreliable Internet coverage (100% of countries), non-payment of data clerks’ incentives (89%), and lack of electronic devices (89%) were the main reasons for the suboptimal functioning of digital systems quoted by key informants. It is critical for investments made and experience acquired in deploying electronic platforms for COVID-19 vaccines to be leveraged to strengthen routine immunization data management.

As the world has become more digitally dependent, questions of data governance, such as ethics, institutional arrangements, and statistical protection measures, have increased in significance. Understanding the economic contribution of investments in data sharing and data governance is highly problematic: outputs and outcomes are often widely dispersed and hard to measure, and the value of those investments is very context-dependent. The “Five Safes” is a popular data governance framework. It is used to design and critique data management strategies across the world and has also been used as a performance framework to measure the effectiveness of data access operations. We report on a novel application of the Five Safes framework to structure the economic evaluation of data governance. The Five Safes was designed to allow structured investigation into data governance. Combining this with more traditional logic models can provide an evaluation methodology that is practical, reproducible, and comparable. We illustrate this by considering the application of the combined logic model-Five Safes framework to data governance for agronomy investments in Ethiopia. We demonstrate how the Five Safes was used to generate the necessary context for a more traditional quantitative study, and consider lessons learned for the wider evaluation of data and data governance investments.