Sparse Historical Data Research Articles

ISASO2: recent trends and regional patterns of ocean dissolved oxygen change

Abstract. Recent estimates of the global inventory of dissolved oxygen (DO) have suggested a decrease of 2 % since the 1960s. However, due to the sparse historical oxygen data coverage, the DO inventory exhibits large regional uncertainties over the interannual timescale. Using the In Situ Analysis System for O2 (ISASO2), a new Argo DO-based optimally interpolated climatology at https://doi.org/10.17882/52367 (Kolodziejczyk et al., 2023), we have estimated an updated regional oxygen inventory. Over the long term (∼ 1980–2013), comparing the ISASO2 Argo fields with the first-guess World Ocean Atlas (WOA18) built from the DO bottle sample fields extracted from the World Ocean Database 2018 (WOD18), the broad tendency to global ocean deoxygenation remains robust in the upper 2000 m, with −451 ± 243 Tmol per decade. The oxygen decline is more pronounced in the key ventilation areas of the Southern Ocean and North Atlantic, except in the Nordic Seas, where oxygen has increased. Over the shorter timescale of the Argo period (2005–2019), the deoxygenation tendency seems globally amplified (−1211 ± 218 Tmol per decade). However, DO changes exhibit stronger amplitude and contrasting regional patterns. The recent changes in Apparent Oxygen Utilization mainly explain the interannual variability in the ventilation regions. However, Argo DO coverage is still incomplete as global and calibration method development is still in progress. Continuing the monitoring of the seasonal-to-interannual and regional-to-global DO variability from ISASO2 will improve our ability to reduce uncertainties in global and regional DO inventories.

Reconstructing the sound speed profile of South China Sea using remote sensing data and long short-term memory neural networks

IntroductionSound waves are refracted along the direction of their propagation owing to spatial and temporal fluctuations in the speed of sound in seawater. Errors are compounded when sound speed profiles (SSPs) with low precision are used to detect and locate distant underwater targets because an accurate SSP is critical for the identification of underwater objects based on acoustic data. Only sparse historical spatiotemporal data on the SSP of the South China Sea are available owing to political issues, its complex atmospheric system, and the unique topography of its seabed, because of which frequent oceanic movements at the mesoscale affect the accuracy of inversion of its SSP.MethodIn this study, we propose a method for the inversion of the SSP of the South China Sea based on a long short-term memory model. We use continuous-time data on the SSP of the South China Sea as well as satellite observations of the height and temperature of the sea surface to make use of the long-term and short-term memory-related capacities of the proposed model.ResultIt can achieve highly accurate results while using a small number of samples by virtue of the unique structure of its memory. Compared with the single empirical orthogonal function regression method, the inversion accuracy of this model is improved by 24.5%, and it performed exceptionally well in regions with frequent mesoscale movements.DiscussionThis enables it to effectively address the challenges posed by the sparse sample distribution and the frequent mesoscale movements of the South China Sea.

Incipient Salinization: A Case Study of the Spring of Asclepieion in Lentas (Ancient Lebena), Crete

Sanctuaries devoted to Asclepius were established and operated for almost a thousand years in various Greek and Roman cities throughout the Mediterranean region. The Asclepieion sanctuary in Lentas (formerly known as Ancient Lebena) in Crete was famous for receiving water from a sacred spring. In Ancient Lebena, Levinaion was a famous centre for hydrotherapy, physiotherapy, and a psychiatric hospital. In the present paper, we aim to assess the hydrochemical status of this sacred spring that holds a prominent position in archaeological and historical studies. The main objectives of this study are: Initially, to present supervisory evidence (archaeological, geological, hydrochemical) of an area that was a water resource management model for many centuries, carrying out therapeutic work. The second objective is to present and compare hydrochemical data in the last century, i.e., from 1915 to 2021. The third objective is to highlight and warn of an incipient saltwater intrusion in the area along the Lentas coast. The fourth objective is to propose an alternative and sustainable form of water resources management in the region that requires the study and rational utilization of the sporadic small water springs in the region. Our study focuses on a basic hydrochemical analysis of spring and borehole water in the remains of Levinaion in the Lentas region, and their comparison with sparse historical data of the sacred spring water, aiming to interpret the impact of the changes in the spring water resources that occurred in recent decades due to urban modernization. Our results highlight (i) visible fluctuations in chemical composition of borehole water samples; (ii) a neutral to alkaline pH in borehole waters and an alkaline pH in spring waters; (iii) undetectable arsenic in Lentas borehole water, unlike historical data of Lentas spring water; (iv) low values of dissolved radon in Lentas borehole water and the spring water of Kefalovrysa; and (v) a timeless constant and hypothermic nature of the water of both the sacred spring and borehole of Lentas, and also of the Kefalovrysa spring. The recorded historical data, i.e., from 1915 to 1957, due to the absence of substantial anthropogenic activity in the area, can be used as reference values (natural background levels, NBLs) for the Lentas area. Our findings emerge with the need to bring again the flowing spring water of the sacred spring of Lentas in its original form through sustainable management and re-discover its beneficial therapeutical effects.

Bayesian averaging-enabled transfer learning method for probabilistic wind power forecasting of newly built wind farms

This paper proposes a technique for the probabilistic wind power forecasting (WPF) of a newly built wind farm (NWF) using a limited amount of historical data. First, the state-of-the-art Transformer network is employed to capture the power generation pattern of different wind farms (WFs) based on abundant historical training samples. Then, the Bayesian averaging regression method is applied to transfer the learned power generation pattern to the NWF by assigning proper weights to the WPF results of different WFs. This enables the proposed method to yield accurate NWF power predictions utilizing a limited amount of historical data. The Bayesian characteristics further enable the quantification of multiple uncertainties in forecasting results that may be essential for the NWF operator when the input is uncertain. Comprehensive tests were also performed by employing other deterministic and probabilistic WPF methods using field data. By comparing the results, the proposed method is demonstrated to produce accurate forecasting results with sparse historical data. Moreover, the uncertainties of outcomes are quantified, and acceptable performance is achieved.

A Quantitative Risk Assessment Framework for the Cybersecurity of Networked Medical Devices

Medical devices are increasingly the source of cybersecurity exposure in healthcare organizations. Research and media reports demonstrate that the exploitation of cybersecurity vulnerabilities can have significant adverse impacts ranging from the exposure of sensitive and personally identifiable patient information to compromising the integrity and availability of clinical care. The results can include identity theft and negative health consequences, including loss of life. Assessing the risk posed by medical devices can provide healthcare organizations with information to prioritize mitigation efforts. However, producing accurate risk assessments in environments with both sparse historical data and a lack of validation regarding the accuracy of forecasts is particularly challenging.
 We present a risk assessment framework for quantifying the risk posed by connected medical devices in trusted healthcare networks. Our framework is built upon prominent existing frameworks and guidance for general risk assessment and cybersecurity risk assessment. We add a method for quantifying risk, which to our knowledge is novel in the context of medical devices on trusted networks. The framework provides a structure for combining publicly available information along with expert elicitation about threats, vulnerabilities, and consequences. The goal is to provide healthcare organizations with actionable information for prioritizing and mitigating risks in medical devices.

Sensitivity of Observationally Based Estimates of Ocean Heat Content and Thermal Expansion to Vertical Interpolation Schemes

AbstractChanges in ocean heat content are a critical element of climate change, with the oceans containing about 90% of the excess heat stored in the climate system and 60% in the upper 700 db. Estimates of these changes are sensitive to horizontal mapping of the sparse historical data and errors in eXpendable BathyThermograph data. Here we show that they are also sensitive to the vertical interpolation of sparsely sampled data through the water column. We estimate, using carefully constructed vertical interpolation methods with high‐quality hydrographic (bottle and CTD) data, the observationally based upper ocean heat content increase (thermosteric sea level rise) from 1956 to 2020 is 285 Zeta Joules (0.55 mm yr−1), 14% (14%) larger than estimates relying on simple but biased linear interpolation schemes. The underestimates have a clear spatial pattern with their maximum near 15°N and 12°S, around the maxima in the curvature of the temperature‐depth profile.

Modelling and predicting enterprise-level cyber risks in the context of sparse data availability

Despite growing attention to cyber risks in research and practice, quantitative cyber risk assessments remain limited, mainly due to a lack of reliable data. This analysis leverages sparse historical data to quantify the financial impact of cyber incidents at the enterprise level. For this purpose, an operational risk database—which has not been previously used in cyber research—was examined to model and predict the likelihood, severity and time dependence of a company’s cyber risk exposure. The proposed model can predict a negative time correlation, indicating that individual cyber exposure is increasing if no cyber loss has been reported in previous years, and vice versa. The results suggest that the probability of a cyber incident correlates with the subindustry, with the insurance sector being particularly exposed. The predicted financial losses from a cyber incident are less extreme than cited in recent investigations. The study confirms that cyber risks are heavy-tailed, jeopardising business operations and profitability.

Aggregating human judgment probabilistic predictions of the safety, efficacy, and timing of a COVID-19 vaccine

Safe, efficacious vaccines were developed to reduce the transmission of SARS-CoV-2 during the COVID-19 pandemic. But in the middle of 2020, vaccine effectiveness, safety, and the timeline for when a vaccine would be approved and distributed to the public was uncertain. To support public health decision making, we solicited trained forecasters and experts in vaccinology and infectious disease to provide monthly probabilistic predictions from July to September of 2020 of the efficacy, safety, timing, and delivery of a COVID-19 vaccine. We found, that despite sparse historical data, a linear pool—a combination of human judgment probabilistic predictions—can quantify the uncertainty in clinical significance and timing of a potential vaccine. The linear pool underestimated how fast a therapy would show a survival benefit and the high efficacy of approved COVID-19 vaccines. However, the linear pool did make an accurate prediction for when a vaccine would be approved by the FDA. Compared to individual forecasters, the linear pool was consistently above the median of the most accurate forecasts. A linear pool is a fast and versatile method to build probabilistic predictions of a developing vaccine that is robust to poor individual predictions. Though experts and trained forecasters did underestimate the speed of development and the high efficacy of a SARS-CoV-2 vaccine, linear pool predictions can improve situational awareness for public health officials and for the public make clearer the risks, rewards, and timing of a vaccine.

Location-Based Web Service QoS Prediction via Preference Propagation to Address Cold Start Problem

Many web-based software systems have been developed in the form of composite services. It is important to accurately predict the Quality of Service (QoS) value of atomic web services because the performance of such composite services depends greatly on the performance of the atomic web service adopted. In recent years, collaborative filtering based methods for predicting the web service QoS values have been proposed. However, they are mainly faced with a cold start problem that is difficult to make reliable prediction due to highly sparse historical data, newly introduced users and web services, and the existing work only deals with the case of newly introduced users. In this article, we propose a Location-based Matrix Factorization using a Preference Propagation method (LMF-PP) to address the cold start problem. LMF-PP fuses invocation and neighborhood similarity, and then the fused similarity is utilized by preference propagation. LMF-PP is compared with existing approaches on the real world dataset. Based on the experimental results, LMF-PP shows better performance than existing approaches in cold start environments as well as in warm start environments.

Damaged Speleothems and Collapsed Karst Chambers Indicate Paleoseismicity of the NE Bohemian Massif (Niedźwiedzia Cave, Poland)

AbstractMultiphase speleothem damage and passage collapse in Niedźwiedzia Cave (NE Bohemian Massif, Poland) were dated with U‐series methods, revealing five events: (1) 320–306 ka, (2) 253–236 ka, (3) 162–158 ka, (4) 132–135 ka, and (5) >21 ka. Events 1, 3 and 4 are robustly constrained, and events 2 and 5 are less certain. Although we cannot unambiguously exclude other agents (frost or gravity collapses), the most likely trigger of damage in the cave was an earthquake, which is supported by timing (the damage occurred independently from climatic conditions in cold and warm periods) and deformation style (damage to the ceiling and walls as well as the passage floor). We applied ground motion models to determine the probable seismic source size, which is most likely the Sudetic Marginal Fault ‐ one of the most pronounced tectonic structures in Central Europe. Located <20 km from the cave and with documented earthquakes of M > 6, the Sudetic Marginal Fault can produce peak ground acceleration values high enough to break speleothems. The other plausible seismic sources are faults in the Upper Nysa Kłodzka Graben located to the east and the Trzebieszowice‐Biela Fault. Although there are sparse historical data that would allow estimating linked seismic hazards, the <8 km distance between the cave and faults should suffice to destroy the speleothems. Niedźwiedzia Cave shielded environmental earthquake effects from erosion. This study shows the advantage of employing speleoseismology in moderate seismic regions, where earthquake effects are rarely preserved in the geological record.

Fitness-for-Service Assessment Approach for Ageing Pipeline Section Based on Sparse Historical Data

Fitness-for-service (FFS) assessment is a common evaluation methodology in oil and gas industries to assess the integrity of in-service structures that may contain flaws, metal thinning and pitting damage. However, given the level of unknowns or missing information in the industry deterministic predictions are unacceptable and invariably the lower bound values could also be substantially conservative. The aim of this work is to develop a generic process to ensure, within a level of confidence, the operational safety and integrity of aging gas or oil pipelines sections based on available data. Fitness for service procedure according to “API 579-1/ASME FFS-1” is performed using local metal loss and micro-cracking to predict a range of safe life for the ageing pipeline operated for around 40 years. The mean value predictions of the present assessment indicate that the flaws away from the weld are within an acceptable boundary which implies the pipes would be fit to continue in operation and at least have 10 years remaining life, whilst the flaws near the weld need to be repaired as soon as possible. This is based on the best average values for the NDE and material property results available. However, adopting extreme caution in the analysis will render the pipes obsolete and ready for replacement. Understanding the risks to be taken in such situations becomes an expert system decision based not just on the FFS analysis but on both quantitative historical data, loading history, material degradation due to environment, corrosion rates and metallurgical analysis in addition to qualitative experience collected from other databases and pipes failure data. Beyond such a procedure the safe option would be a full burst pressure testing of the length of pipeline in question to identify possible leaks of old pipes.

DP-BPR: Destination prediction based on Bayesian personalized ranking

Destination prediction has attracted widespread attention because it can help vehicle-aid systems recommend related services in advance to improve user driving experience. However, the relevant research is mainly based on driving trajectory of vehicles to predict the destinations, which is challenging to achieve the early destination prediction. To this end, we propose a model of early destination prediction, DP-BPR, to predict the destinations by users’ travel time and locations. There are three challenges to accomplish the model: 1) the extremely sparse historical data make it challenge to predict destinations directly from raw historical data; 2) the destinations are related to not only departure points but also departure time so that both of them should be taken into consideration in prediction; 3) how to learn destination preferences from historical data. To deal with these challenges, we map sparse high-dimensional data to a dense low-dimensional space through embedding learning using deep neural networks. We learn the embeddings not only for users but also for locations and time under the supervision of historical data, and then use Bayesian personalized ranking (BPR) to learn to rank destinations. Experimental results on the Zebra dataset show the effectiveness of DP-BPR.

Changes in the environmental impacts of pig production systems in Great Britain over the last 18 years

ContextThe aims of global pig production systems include a reduction of their environmental impacts, which can be achieved through an increase in outputs whilst minimising inputs. ObjectiveThe aim of this paper was to develop a novel method to enable estimation of the changes in the environmental impacts based on sparse data from the British pig production industry over ~20 years. MethodsTo achieve this, we developed a Life Cycle Assessment (LCA) method capable of dealing with sparse historical data from livestock systems. We applied it, for the first time, to estimate the temporal changes in environmental impacts of British pig production systems caused by changes in production performance. Performance data available from industry-held databases for indoor and outdoor bred pigs in Britain were used to estimate nutrient requirements through animal performance modelling, and feed composition through least-cost formulation. The cradle to farm-gate LCA model developed, included manure management and the full life cycle of the pigs and its functional unit was 1 kg of live weight pig at farm-gate. Sensitivity analyses were conducted to investigate the potential influence of changes in animal performance and feed prices on the estimated changes in environmental impacts. Results and ConclusionsThe higher growth rates and increased leanness over the period considered led to substantial reductions in energy requirements. Overall, the system changes led to reductions, for indoor and outdoor bred systems respectively, of 37.0% and 35.4% for Global Warming Potential, 21.2% and 16.4% in Terrestrial Acidification Potential, 22.5% and 22.3% in Freshwater Eutrophication Potential, 15.8% and 16.8% in Agricultural Land Use and 16.5% and 16.1% in Fossil Resource Scarcity. The sensitivity analyses showed that trends in feed composition were influential on the environmental impact outcomes, and that the LCA model was more sensitive to the change in feed composition than to the changes in animal performance over the time period considered. SignificanceKnowledge of temporal changes to the environmental impacts of livestock systems and the drivers of changes to date should guide future decisions to mitigate these impacts.

Deep Learning-Enabled Sparse Industrial Crowdsensing and Prediction

Mobile Crowdsensing (MCS) is a powerful sensing paradigm, which provides sufficient social data for cognitive analytics in industrial sensing, and industrial manufacturing. Considering the sensing costs, sparse MCS, as a variant, only senses the data in a few subareas, and then infers the data of unsensed subareas by the spatio-temporal relationship of the sensed data. Existing works usually assume that the sensed data are linearly spatiotemporal dependent, which cannot work well in real-world nonlinear systems, and thus, result in low data inference accuracy. Moreover, in many cases, users not only require inferring the current data, but also have an interest in predicting the near future, which can provide more information for users' decision making. Facing these problems, we propose a deep learning-enabled industrial sensing, and prediction scheme based on sparse MCS, which consists of two parts: matrix completion and future prediction. Our goal is to achieve high-precision prediction of future moments under the hypothesis of sparse historical data. To make full use of the sparse data for prediction, we first propose a deep matrix factorization method, which can retain the nonlinear temporal-spatial relationship, and perform high-precision matrix completion. In order to predict the subareas' data in several future sensing cycles, we further propose a nonlinear autoregressive neural network, and a stacked denoising autoencoder to obtain the temporal-spatial correlation between the data from different cycles or subareas. According to the results gained by experiments on four real-world industrial sensing datasets consisting of six typical tasks, it can be seen that the method in this article improves the accuracy of prediction using sparse data.

The Saffarid Yaʿqub b. Layth and the Revival of Persian Kingship

Abstract The sparse historical data suggest that Yaʿqub b. Layth was the first Iranian ruler following the Arab-Islamic conquests to make significant efforts to revive Persian kingship. This article seeks to clarify, as far as possible, Yaʿqub’s actions and goals, as well as the context for his efforts. This interpretation of the sources argues that the Saffarid ruler’s government faced a crisis of social legitimacy owing to the tense relations that developed between Yaʿqub and the caliph on the one hand, and the nobles of Khorasan on the other. In this context, only in the last four years of his rule was Yaʿqub, in an effort to legitimate his power, forced to turn away from dependence on caliphal investiture and to appeal instead to a revival of Persian kingship (without the official use of the title “king”).

Modelling global tropical cyclone wind footprints

Abstract. A novel approach to modelling the surface wind field of landfalling tropical cyclones (TCs) is presented. The modelling system simulates the evolution of the low-level wind fields of landfalling TCs, accounting for terrain effects. A two-step process models the gradient-level wind field using a parametric wind field model fitted to TC track data and then brings the winds down to the surface using a numerical boundary layer model. The physical wind response to variable surface drag and terrain height produces substantial local modifications to the smooth wind field provided by the parametric wind profile model. For a set of US historical landfalling TCs the accuracy of the simulated footprints compares favourably with contemporary modelling approaches. The model is applicable from single-event simulation to the generation of global catalogues. One application demonstrated here is the creation of a dataset of 714 global historical TC overland wind footprints. A preliminary analysis of this dataset shows regional variability in the inland wind speed decay rates and evidence of a strong influence of regional orography. This dataset can be used to advance our understanding of overland wind risk in regions of complex terrain and support wind risk assessments in regions of sparse historical data.

Effect of Epicutaneous Immunotherapy Versus Placebo on Reaction to Peanut Protein Ingestion Among Children With Peanut Allergy: The PEPITES Randomized Clinical Trial

To assess the efficacy and adverse events of epicutaneous immunotherapy with a peanut patch among children with peanut allergy.The study included 356 children with peanut allergy aged 4 to 11 years without a history of severe peanut-related anaphylaxis who showed objective signs and symptoms of reactivity at an eliciting dose of ≤300 mg of peanut protein during a double-blinded, placebo-controlled food challenge.This study was a phase 3, randomized, double-blinded, placebo-controlled multicenter trial. Study participants were stratified into low- (≤10 mg) and high-eliciting dose (10–300 mg) subgroups on the basis of participants’ screening eliciting dose. Participants were then randomly assigned into groups receiving a daily epicutaneous patch containing 250 µg of peanut protein (n = 238) or a daily placebo patch (n = 118). Treatment response was defined as a posttreatment eliciting dose of ≥300 or ≥1000 mg of peanut protein for the pretreatment low- and high-eliciting dose subgroups, respectively. The treatment effect of the immunotherapy was determined by the response rate difference between the treatment and placebo groups after 12 months. A margin of 15% was selected as the lower bound of the 95% confidence interval (CI) for determining a positive trial result. The incidence of treatment-emergent adverse events (TEAEs) was also followed in each group.After 12 months, the response rates of the treatment (35.3%) and placebo (13.6%) groups were compared, showing a difference of 21.7% that was statistically significant (P < .001). However, the resulting lower bound of the 95% CI was 12.4%, which did not achieve the pretrial 15% criterion for clinically relevant significance. The incidence of TEAEs was 95.4% in the treatment group and 89% in the placebo group, in which application site reactions were the most common TEAE observed in both groups. There were 4 TEAEs in 3 treatment group participants that were classified as moderate anaphylactic reactions, all cases were appropriately treated, and 2 of these participants permanently discontinued the study.This study demonstrated a statistically significant difference between response rates of the peanut patch and placebo patch groups on the basis of eliciting doses during food challenges before and after the 12-month trial period. The results could not be considered a positive trial because the lower bound of the 95% CI was not met.Epicutaneous immunotherapy stands as a candidate for therapy for children with peanut allergy who would benefit from desensitization and lower risk of potentially life-threatening anaphylaxis resulting from accidental peanut exposure. Given sparse historical data for the assessment of food allergy immunotherapy trials, the lower bound criterion for the 95% CI in this study was set at 15% under US Food and Drug Administration guidance but is considered by some to be arbitrary in nature. The failure of this study’s results to meet this preset criterion signifies the results may not be clinically significant, despite showing a statistically significant difference in response rate between the treatment and placebo groups. Nevertheless, this treatment option may offer some clinical benefit to a vulnerable population, but this should be weighed on the generally low incidence of severe reactions in light of appropriate food avoidance. A longer trial, beyond the 12-month duration of this study, may prove to show clinically relevant results and may offer further insight into the maintenance of desensitization.

Genetic Analysis in the Tanzania Sickle Surveillance Study (TS3): Modifiers of Sickle Cell Disease and Identification of Hemoglobin Variants

Introduction. Based on sparse historical data, Tanzania ranks fourth globally for the estimated number of annual births with sickle cell disease. Northwest Tanzania is projected to have an especially high burden of sickle cell disease, but no contemporary surveillance data exist to verify this projection. We designed a large prospective study to determine the prevalence of both sickle cell trait and sickle cell disease in northwest Tanzania. Additional objectives included an analysis of two known genetic modifiers of sickle cell disease [alpha thalassemia trait and glucose-6-phosphate dehydrogenase (G6PD) deficiency], as well as genetic variants affecting fetal hemoglobin (HbF) expression and characterization of hemoglobin variants. Methods. The Tanzania Sickle Surveillance Study (TS3) is a prospective cross-sectional study to determine the prevalence of sickle cell trait and disease in an area projected to be at high risk. Dried blood spots (DBS) from all children born to HIV-infected mothers in the 9 regions across northwest Tanzania were collected by the HIV Early Infant Diagnosis (EID) program and transported to a central laboratory at Bugando Medical Centre, a teaching and consultancy hospital in Mwanza. DBS were tested by isoelectric focusing and scored as normal, trait, disease, or variant. Samples scored as disease or variant were retested for confirmation, and then frozen for subsequent genetic analysis. In the US, genomic DNA was extracted from each DBS, and sickle cell disease status was confirmed by PCR and TaqMan genotyping. All confirmed samples were then analyzed for alpha thalassemia trait (rightward -3.7 gene deletion) using quantitative real time PCR, and for G6PD deficiency using three real time PCR probes to distinguish A and B G6PD isoforms, to identify the G6PD A- variant, and to confirm the sex, respectively. Single nucleotide polymorphisms within 3 quantitative trait loci [BCL11A, the HBS1L-MYB intergenic polymorphism (HMIP) region, and the gamma-globin promoter XmnI site] that modify baseline HbF were analyzed using TaqMan genotyping. DBS with hemoglobin variants were investigated for the presence of common variants found in East Africa. Results. Between February 2017 and May 2018, a total of 232 IEF gels were completed by local staff in Tanzania. Among EID samples from children &amp;lt;24 months of age, the median age was 52 days (IQR 41-93 days), and 17,278 unique DBS samples were scored. With 20 samples uninterpretable and 54 samples with missing results, the primary analysis was performed on 17,204 samples. The overall prevalence of sickle cell trait and disease in the entire cohort was 20.3% and 1.2%, respectively, for a sickle allele frequency of 0.114 with Hardy-Weinberg equilibrium. There was some geographical variation between individual districts in both sickle trait (15.2-27.8%) and disease (0.0-4.3%). Hemoglobin variants were rare (0.1%) and included 4 Hb G-Pest, 2 Hb Kenya, and 1 Hb P-Nilotic. Of 143 DBS confirmed to have sickle cell disease and available for further testing, 61 (43%) had one-gene deletion alpha thalassemia trait and 21 (15%) had two-gene deletion. A- G6PD deficiency was detected in 19.2% of males, and 25.7% of females were heterozygous carriers. The minor allelic frequency for known modifiers of HbF were 32.2% C at rs11886868 (BCL11A), 26.6% T at rs1427407 (BCL11A), 32.5% A at rs4671393 (BCL11A), 23.8% G at rs28384513 (HMIP), 4.5% C at rs9399137 (HMIP), and 0% A at rs7482144 (XmnI). Conclusion. The prevalence of sickle cell trait and disease among infants in northwest Tanzania is very high, exceeding 20% trait and 1.2% disease. Regional prevalence data paired with region-specific crude birth rates predict 10,055 births annually in the northwest regions, more than doubling previous estimates. Concomitant alpha thalassemia trait and G6PD deficiency are frequently co-inherited and may affect the phenotype, as well as common genetic modifiers of HbF expression. Our detailed genetic analysis of a geographically representative surveillance cohort provides a foundation for future targeted screening and the introduction of hydroxyurea for treatment of sickle cell disease in northwest Tanzania. Disclosures Ware: Bristol Myers Squibb: Other: Research Drug Donation; Addmedica: Other: Research Drug Donation; Global Blood Therapeutics: Membership on an entity's Board of Directors or advisory committees; Agios: Membership on an entity's Board of Directors or advisory committees; Nova Laboratories: Membership on an entity's Board of Directors or advisory committees; CSL Behring: Membership on an entity's Board of Directors or advisory committees; Novartis: Other: DSMB.

A Personal Location Prediction Method Based on Individual Trajectory and Group Trajectory

With the rapid development of communication technology, a large amount of spatiotemporal trajectory data has been produced. One of the critical applications of trajectory data is location prediction that is important for urban traffic planning and location-based services. Although many methods for personal location prediction have been proposed, to the best of our knowledge, some users always have sparse historical trajectory data in practical applications, resulting in poor location prediction precision. Targeting on this challenge, we propose an Individual Trajectory-Group Trajectory (ITGT) location prediction model by utilizing the pattern of group travels. First, the model performs the stay point extraction and conducts the spatial clustering to construct the clustering link. Second, Fano's inequality and clustering link are used to evaluate the predictability of location. Third, two variable order Markov models, named prediction by partial match (PPM) and probabilistic suffix tree (PST), are adopted to predict the clustering link. Finally, our approach is evaluated by using 608 712 points from 5000 volunteers at Shenzhen, China. The results show that: 1) when using individual trajectory, the PPM individual model is superior to the conventional N-order Markov model and PST individual model; 2) when all group trajectories are used, the PPM group model is not as accurate as the PPM individual model; and 3) when it classifies the group trajectory into different traffic zones, the PPM zone model is better than the PPM group model and the PPM individual model. The prediction precision of 1-3 order PPM zone model is 83.21%, 86.75%, and 87.35%, respectively, which introduces approximately 3% performance gains by utilizing the characters of traffic zone groups.

Social-Aware Video Recommendation for Online Social Groups

Group recommendation plays a significant role in today's social media systems, where users form social groups to receive multimedia content together and interact with each other, instead of consuming the online content individually. Limitations of traditional group recommendation approaches are as follows. First, they usually infer group members’ preferences by their historical behaviors, failing to capture inactive users’ preferences from the sparse historical data. Second, relationships between group members are not studied by these approaches, which fail to capture the inherent personality of members in a group. To address these issues, we propose a social-aware group recommendation framework that jointly utilizes both social relationships and social behaviors to not only infer a group's preference, but also model the tolerance and altruism characteristics of group members. Based on the observation that the following relationship in the online social network reflects common interests of users, we propose a group preference model based on external experts of group members. Furthermore, we model users’ tolerance (willingness to receive content not preferred) and altruism (willingness to receive content preferred by friends). Finally, based on the group preference model, we design recommendation algorithms for users under different social contexts. Experimental results demonstrate the effectiveness of our approach, which significantly improves the recommendation accuracy against traditional approaches, especially in the cases of inactive group members.