This paper presents a new two-step approach for record linkage, focusing on the creation of high-quality training data in the first step. The approach employs the unsupervised random forest model as a similarity measure to produce a similarity score vector for record matching. Three constructions were proposed to select non-match pairs for the training data, with both balanced (symmetry) and imbalanced (asymmetry) distributions tested. The top and imbalanced construction was found to be the most effective in producing training data with 100% correct labels. Random forest and support vector machine classification algorithms were compared, and random forest with the top and imbalanced construction produced an F₁-score comparable to probabilistic record linkage using the expectation maximisation algorithm and EpiLink. On average, the proposed approach using random forests and the top and imbalanced construction improved the F₁-score by 1% and recall by 6.45% compared to existing record linkage methods. By emphasising the creation of high-quality training data, this new approach has the potential to improve the accuracy and efficiency of record linkage for a wide range of applications.

Malaysia is multi-ethnic and diverse country. Heterogeneity, in terms of population interactions, is ingrained in the foundation of the country. Malaysian policies and social distancing measures are based on daily infections and R0 (average number of infections per infected person), estimated from the data. Models of the Malaysian COVID-19 spread are mostly based on the established SIR compartmental model and its variants. These models usually assume homogeneity and symmetrical full mixing in the population; thus, they are unable to capture super-spreading events which naturally occur due to heterogeneity. Moreover, studies have shown that when heterogeneity is present, R0 may be very different and even possibly misleading. The underlying spreading network is a crucial element, as it introduces heterogeneity for a more representative and realistic model of the spread through specific populations. Heterogeneity introduces more complexities in the modelling due to its asymmetrical nature of infection compared to the relatively symmetrical SIR compartmental model. This leads to a different way of calculating R0 and defining super-spreaders. Quantifying a super-spreader individual is related to the idea of importance in a network. The definition of a super-spreading individual depends on how super-spreading is defined. Even when the spreading is defined, it may not be clear that a single centrality always correlates with super-spreading, since centralities are network dependent. We proposed using a measure of super-spreading directly related to R0 and that will give a measure of ‘spreading’ regardless of the underlying network. We captured the vulnerability for varying degrees of heterogeneity and initial conditions by defining a measure to quantify the chances of epidemic spread in the simulations. We simulated the SIR spread on a real Malaysian network to illustrate the effects of this measure and heterogeneity on the number of infections. We also simulated super-spreading events (based on our definition) within the bounds of heterogeneity to demonstrate the effectiveness of the newly defined measure. We found that heterogeneity serves as a natural curve-flattening mechanism; therefore, the number of infections and R0 may be lower than expected. This may lead to a false sense of security, especially since heterogeneity makes the population vulnerable to super-spreading events.

The widely used Poisson count process in insurance claims modeling is no longer valid if the claims occurrences exhibit dispersion. In this paper, we consider the aggregate discounted claims of an insurance risk portfolio under Weibull counting process to allow for dispersed datasets. A copula is used to define the dependence structure between the interwaiting time and its subsequent claims amount. We use a Monte Carlo simulation to compute the higher-order moments of the risk portfolio, the premiums and the value-at-risk based on the New Zealand catastrophe historical data. The simulation outcomes under the negative dependence parameter $\theta$, shows the highest value of moments when claims experience exhibit overdispersion. Conversely, the underdispersed scenario yields the highest value of moments when $\theta$ is positive. These results lead to higher premiums being charged and more capital requirements to be set aside to cope with unfavorable events borne by the insurers.

Coronavirus COVID-19 shocking the whole world due to its highly contagious characteristics implicating not only public health, but also economy and social life. Since the effects are momentous, plenty of research have been conducted and still ongoing in order to study and to learn more about this virus and how it changing our daily life. In this paper, we explore 134 articles published in 2020 related to COVID-19 and narrowing the scope of study to Malaysia. An alternative route was taken by employing Latent Dirichlet Allocation (LDA) to identify underlying themes or topics in these publications. Two separate analyses were conducted, one is to the paper’s titles and another one to the journal’s names. The findings identified three topics for paper’s titles data are clinical study, impact of COVID-19 on various fields and Movement Control Order (MCO). The last topic shows the locality criterion in the studied papers as the term MCO was only used in Malaysia. For the journal’s names, three topics identified were medical study, public health also business and education. Two papers with the most number of citations are both in social sciences. Investigating the properties of these topics, we found that papers on clinical studies are the ones with more chance to be cited and published by reputable publishers. These findings may help researchers on planning and strategizing for future research on COVID-19 specifying on Malaysia cases.