The Poisson equation is commonly encountered in engineering, for instance, in computational fluid dynamics (CFD) where it is needed to compute corrections to the pressure field to ensure the incompressibility of the velocity field. In the present work, we propose a novel fully convolutional neural network (CNN) architecture to infer the solution of the Poisson equation on a 2D Cartesian grid with different resolutions given the right-hand side term, arbitrary boundary conditions, and grid parameters. It provides unprecedented versatility for a CNN approach dealing with partial differential equations. The boundary conditions are handled using a novel approach by decomposing the original Poisson problem into a homogeneous Poisson problem plus four inhomogeneous Laplace subproblems. The model is trained using a novel loss function approximating the continuous $ {L}^p $ norm between the prediction and the target. Even when predicting on grids denser than previously encountered, our model demonstrates encouraging capacity to reproduce the correct solution profile. The proposed model, which outperforms well-known neural network models, can be included in a CFD solver to help with solving the Poisson equation. Analytical test cases indicate that our CNN architecture is capable of predicting the correct solution of a Poisson problem with mean percentage errors below 10%, an improvement by comparison to the first step of conventional iterative methods. Predictions from our model, used as the initial guess to iterative algorithms like Multigrid, can reduce the root mean square error after a single iteration by more than 90% compared to a zero initial guess.

Monte Carlo algorithms simulates some prescribed number of samples, taking some random real time to complete the computations necessary. This work considers the converse: to impose a real-time budget on the computation, which results in the number of samples simulated being random. To complicate matters, the real time taken for each simulation may depend on the sample produced, so that the samples themselves are not independent of their number, and a length bias with respect to compute time is apparent. This is especially problematic when a Markov chain Monte Carlo (MCMC) algorithm is used and the final state of the Markov chain—rather than an average over all states—is required, which is the case in parallel tempering implementations of MCMC. The length bias does not diminish with the compute budget in this case. It also occurs in sequential Monte Carlo (SMC) algorithms, which is the focus of this paper. We propose an anytime framework to address the concern, using a continuous-time Markov jump process to study the progress of the computation in real time. We first show that for any MCMC algorithm, the length bias of the final state’s distribution due to the imposed real-time computing budget can be eliminated by using a multiple chain construction. The utility of this construction is then demonstrated on a large-scale SMC$ {}^2 $ implementation, using four billion particles distributed across a cluster of 128 graphics processing units on the Amazon EC2 service. The anytime framework imposes a real-time budget on the MCMC move steps within the SMC$ {}^2 $ algorithm, ensuring that all processors are simultaneously ready for the resampling step, demonstrably reducing idleness to due waiting times and providing substantial control over the total compute budget.

The COVID-19 pandemic is a global challenge for humanity, in which a large number of resources are invested to develop effective vaccines and treatments. At the same time, governments try to manage the spread of the disease while alleviating the strong impact derived from the slowdown in economic activity. Governments were forced to impose strict lockdown measures to tackle the pandemic. This significantly changed people’s mobility and habits, subsequently impacting the economy. In this context, the availability of tools to effectively monitor and quantify mobility was key for public institutions to decide which policies to implement and for how long. Telefonica has promoted different initiatives to offer governments mobility insights throughout many of the countries where it operates in Europe and Latin America. Mobility indicators with high spatial granularity and frequency of updates were successfully deployed in different formats. However, Telefonica faced many challenges (not only technical) to put these tools into service in a short timing: from reducing latency in insights to ensuring the security and privacy of information. In this article, we provide details on how Telefonica engaged with governments and other stakeholders in different countries as a response to the pandemic. We also cover the challenges faced and the shared learnings from Telefonica’s experience in those countries.

The rapid spread of COVID-19 infections on a global level has highlighted the need for accurate, transparent and timely information regarding collective mobility patterns to inform de-escalation strategies as well as to provide forecasting capacity for re-escalation policies aiming at addressing further waves of the virus. Such information can be extracted using aggregate anonymized data from innovative sources such as mobile positioning data. This paper presents lessons learnt and results of a unique Business-to-Government initiative between several mobile network operators in Europe and the European Commission. Mobile positioning data have supported policy-makers and practitioners with evidence and data-driven knowledge to understand and predict the spread of the disease, the effectiveness of the containment measures, their socio-economic impacts while feeding scenarios at European Union scale and in a comparable way across countries. The challenges of these data sharing initiative are not limited to data quality, harmonization, and comparability across countries, however important they are. Equally essential aspects that need to be addressed from the onset are related to data privacy, security, fundamental rights, and commercial sensitivity.

An outbreak surveillance system for Salmonella integrating whole genome sequencing (WGS) and epidemiological data was developed in South East and London in 2016–17 to assess local WGS clusters for triage and investigation. Cases genetically linked within a 5 single-nucleotide polymorphism (SNP) single linkage cluster were assessed using a set of locally agreed thresholds based on time, person and place, for reporting to local health protection teams (HPTs). Between September 2016 and September 2017, 230 unique 5-SNP clusters (442 weekly reports) of non-typhoidal Salmonella 5-SNP WGS clusters were identified, of which 208 unique 5-SNP clusters (316 weekly reports) were not reported to the HPTs. In the remaining 22 unique clusters (126 weekly clusters) reported to HPTs, nine were known active outbreak investigations, seven were below locally agreed thresholds and six exceeded local thresholds. A common source or vehicle was identified in four of six clusters that exceeded locally agreed thresholds. This work demonstrates that a threshold-based surveillance system, taking into account time, place and genetic relatedness, is feasible and effective in directing the use of local public health resources for risk assessment and investigation of non-typhoidal Salmonella clusters.

Coxiella burnetii is a zoonotic agent responsible for human Q fever, a potentially severe disease that can lead to persistent infection. This cross-sectional study aimed to estimate the seroprevalence to C. burnetii antibodies and its association with potential risk factors in the human population of five regions of Québec, Canada. A serum bank comprising sera from 474 dog owners was screened by an enzyme-linked immunosorbent assay followed by confirmation of positive or equivocal sera by an indirect immunofluorescence assay. Observed seroprevalences of 1.2% (95% confidence interval (CI): 0.0–6.6), 2.6% (95% CI: 0.5–7.4) and 5.9% (95% CI: 3.4–9.6) were estimated in the regions of Montréal, Lanaudière and Montérégie, respectively, which all included at least 83 samples. Having lived or worked on a small ruminant farm (prevalence odds ratio (POR) = 5.4; 95% CI: 1.6–17.7) and being a veterinarian or veterinary student (POR = 6.1; 95% CI: 1.6–24.0) were significantly associated with C. burnetii seropositivity. Antibodies against C. burnetii were detected in the human population of Québec. Although seropositivity to this agent was associated with occupational contact with domestic animals, antibodies were also detected in people with no reported professional exposure. No associations with ruminant farm proximity were identified.

Many financial time series have varying structures at different quantile levels, and also exhibit the phenomenon of conditional heteroskedasticity at the same time. However, there is presently no time series model that accommodates both of these features. This paper fills the gap by proposing a novel conditional heteroskedastic model called “quantile double autoregression”. The strict stationarity of the new model is derived, and self-weighted conditional quantile estimation is suggested. Two promising properties of the original double autoregressive model are shown to be preserved. Based on the quantile autocorrelation function and self-weighting concept, three portmanteau tests are constructed to check the adequacy of the fitted conditional quantiles. The finite sample performance of the proposed inferential tools is examined by simulation studies, and the need for use of the new model is further demonstrated by analyzing the S&P500 Index.

Aggregated data from mobile network operators (MNOs) can provide snapshots of population mobility patterns in real time, generating valuable insights when other more traditional data sources are unavailable or out-of-date. The COVID-19 pandemic has highlighted the value of remotely-collected, high-frequency, localized data in inferring the economic impact of shocks to inform decision-making. However, proper protocols must be put in place to ensure end-to-end user-confidentiality and compliance with international best practice. We demonstrate how to build such a data pipeline, channeling data from MNOs through the national regulator to the analytical users, who in turn produce policy-relevant insights. The aggregated indicators analyzed offer a detailed snapshot of the decrease in mobility and increased out-migration from urban to rural areas during the COVID-19 lockdown. Recommendations based on lessons learned from this process can inform engagements with other regulators in creating data pipelines to inform policy-making.

In this paper, we propose two bootstrap procedures, namely parametric and block bootstrap, to approximate the finite sample distribution of change-point estimators for piecewise stationary time series. The bootstrap procedures are then used to develop a generalized likelihood ratio scan method (GLRSM) for multiple change-point inference in piecewise stationary time series, which estimates the number and locations of change-points and provides a confidence interval for each change-point. The computational complexity of using GLRSM for multiple change-point detection is as low as $O(n(\log n)^{3})$ for a series of length n. Extensive simulation studies are provided to demonstrate the effectiveness of the proposed methodology under different scenarios. Applications to financial time series are also illustrated.

Severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) pandemic is still ongoing along with the global vaccination efforts against it. Here, we aimed to understand the longevity and strength of anti-SARS-CoV-2 IgG responses in a small community (n = 283) six months following local SARS-COV-2 outbreak in March 2020. Three serological assays were compared and neutralisation capability was also determined. Overall 16.6% (47/283) of the participants were seropositive and 89.4% (42/47) of the IgG positives had neutralising antibodies. Most of the symptomatic individuals confirmed as polymerase chain reaction (PCR) positive during the outbreak were seropositive (30/32, 93.8%) and 33.3% of the individuals who quarantined with a PCR confirmed patient had antibodies. Serological assays comparison revealed that Architect (Abbott) targeting the N protein LIASON® (DiaSorin) targeting the S protein and enzyme-linked immunosorbent assay (ELISA) targeting receptor binding domain detected 9.5% (27/283), 17.3% (49/283) and 17% (48/283), respectively, as IgG positives. The latter two assays highly agreed (kappa = 0.89) between each other. In addition, 95%, (19/20, by ELISA) and 90.9% (20/22, with LIASON) and only 71.4% (15/21, by Architect) of individuals that were seropositive in May 2020 were found positive also in September. The unexpected low rate of overall immunity indicates the absence of un-noticed, asymptomatic infections. Lack of overall high correlation between the assays is attributed mainly to target-mediated antibody responses and suggests that using a single serological assay may be misleading.

During the COVID-19 crisis, the French National Institute of Statistics and Economic Studies (INSEE) used aggregated and anonymous counting indicators based on network signaling data of three of the four mobile network operators (MNOs) in France to measure the distribution of population over the territory during and after the lockdown and to enrich the toolbox of high-frequency economic indicators used to follow the economic situation. INSEE’s strategy was to combine information coming from different MNOs together with the national population estimates it usually produces in order to get more reliable statistics and to measure uncertainty. This paper relates and situates this initiative within the long-term methodological collaborations between INSEE and different MNOs, and INSEE, Eurostat, and some other European national statistical institutes (NSIs). These collaborations aim at constructing experimental official statistics on the population present in a given place and at a given time, from mobile phone data (MPD). The COVID-19 initiative has confirmed that more methodological investments are needed to increase relevance of and trust in these data. We suggest this methodological work should be done in close collaboration between NSIs, MNOs, and research, to construct the most reliable statistical processes. This work requires exploiting raw data, so the research and statistical exemptions present in the general data protection regulation (GDPR) should be introduced as well in the new e-privacy regulation. We also raise the challenges of articulating commercial and public interest rationales and articulating transparency and commercial secrets requirements. Finally, it elaborates on the role NSIs can play in the MPD valorization ecosystem.

We assessed severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) reverse transcriptase-polymerase chain reaction (RT-PCR) diagnostic sensitivity and cycle threshold (Ct) values relative to symptom onset in symptomatic coronavirus disease-2019 (COVID-19) patients from Bavaria, Germany, of whom a subset was repeatedly tested. Locally weighted scatterplot smoothing method was used to assess the relationship between symptom onset and Ct-values. Kaplan−Meier plots were used to visualise the empirical probability of detecting viral ribonucleic acid (RNA) over time and estimate the time until clearance of viral RNA among the repeatedly tested patients. Among 721 reported COVID-19 cases, the viral RNA was detected in specimens taken between three days before and up to 48 days after symptom onset. The mean Ct-value was 28.6 (95% confidence interval (CI) 28.2–29.0) with the lowest mean Ct-value (26.2) observed two days after symptom onset. Up to 7 days after symptom onset, the diagnostic sensitivity of the RT-PCR among repeatedly sampled patients (n = 208) remained above 90% and decreased to 50% at day 12 (95% CI 10.5–21.5). Our data provide valuable estimates to optimise the timing of sampling of individuals for SARS-CoV-2 detection. A considerable proportion of specimens sampled before symptom onset had Ct-values comparable with Ct-values after symptom onset, suggesting the probability of presymptomatic transmission.

Telecommunications data are being explored by many countries as a new source of data that can be incorporated into their national statistical systems. In particular, “mobile positioning data” are increasingly being used to study population movements and population distributions. However, the legal, ethical, and technical complexities of working with this type of data often pose many barriers, which can prevent the data from being used at the times when it is most urgently needed. We demonstrate how having a robust public–private partnership framework, a privacy-preserving technical setup, and a communications strategy already in place, prior to an emergency, can enable governments to harness the advantages of telecommunications data at the times when it is most valuable. However, even once these foundations are in place, the challenges of competing priorities, managing expectations, and maintaining communication with data consumers during a pandemic mean that the potential of the data is not automatically translated into direct impact. This highlights the importance of sensitisation exercises, targeted at potential data users, to make clear the potential and limitations of the data, as well as the importance of being able to maintain direct communication with data users. The views expressed in this work belong solely to the authors and should not be interpreted as the views of their institutions.

In the absence of an established gold standard, an understanding of the testing cycle from individual exposure to test outcome report is required to guide the correct interpretation of severe acute respiratory syndrome-coronavirus-2 reverse transcriptase real-time polymerase chain reaction (RT-PCR) results and optimise the testing processes. Bayesian network models have been used within healthcare to bring clarity to complex problems. We use this modelling approach to construct a comprehensive framework for understanding the real-world predictive value of individual RT-PCR results.

We elicited knowledge from domain experts to describe the test process through a facilitated group workshop. A preliminary model was derived based on the elicited knowledge, then subsequently refined, parameterised and validated with a second workshop and one-on-one discussions.

Causal relationships elicited describe the interactions of pre-testing, specimen collection and laboratory procedures and RT-PCR platform factors, and their impact on the presence and quantity of virus and thus the test result and its interpretation. By setting the input variables as ‘evidence’ for a given subject and preliminary parameterisation, four scenarios were simulated to demonstrate potential uses of the model.

The core value of this model is a deep understanding of the total testing cycle, bridging the gap between a person's true infection status and their test outcome. This model can be adapted to different settings, testing modalities and pathogens, adding much needed nuance to the interpretations of results.

To investigate the current epidemiology of melioidosis in Yangon, Myanmar, between June 2017 and May 2019 we conducted enhanced surveillance for melioidosis in four tertiary hospitals in Yangon, where the disease was first discovered in 1911. Oxidase-positive Gram-negative rods were obtained from the microbiology laboratories and further analysed at the Department of Medical Research. Analysis included culture on Ashdown agar, the three disc sensitivity test (gentamicin, colistin and co-amoxiclav), latex agglutination, API 20 NE, antibiotic susceptibility testing, and a subset underwent molecular confirmation with a Burkholderia pseudomallei specific assay. Twenty one of 364 isolates (5.7%) were confirmed as B. pseudomallei and were mostly susceptible to the antibiotics used in standard therapy for melioidosis. Ten patients were from Yangon Region, nine were from Ayeyarwaddy region, and one each was from Kayin and Rakhine States. A history of soil contact was given by seven patients, five had diabetes mellitus and one had renal insufficiency. The patients presented with septicaemia (12 cases), pneumonia (three cases), urinary tract infection (two cases) and wound infection (four cases). Eighteen patients survived to hospital discharge. This study highlights the likelihood that melioidosis may be far more common, but underdiagnosed, in more rural parts of Myanmar as in other countries in SE Asia.

Developing agents capable of commonsense reasoning is an important goal in Artificial Intelligence (AI) research. Because commonsense is broadly defined, a computational theory that can formally categorize the various kinds of commonsense knowledge is critical for enabling fundamental research in this area. In a recent book, Gordon and Hobbs described such a categorization, argued to be reasonably complete. However, the theory’s reliability has not been independently evaluated through human annotator judgments. This paper describes such an experimental study, whereby annotations were elicited across a subset of eight foundational categories proposed in the original Gordon-Hobbs theory. We avoid bias by eliciting annotations on 200 sentences from a commonsense benchmark dataset independently developed by an external organization. The results show that, while humans agree on relatively concrete categories like time and space, they disagree on more abstract concepts. The implications of these findings are briefly discussed.