Spatiotemporal Prediction of COVID-19 Mortality and Risk Assessment

This paper presents a multivariate functional data statistical approach, for spatiotemporal prediction of COVID-19 mortality counts. Specifically, spatial heterogeneous nonlinear parametric functional regression trend model fitting is first implemented. Classical and Bayesian infinite-dimensional log-Gaussian linear residual correlation analysis is then applied. The nonlinear regression predictor of the mortality risk is combined with the plug-in predictor of the multiplicative error term. An empirical model ranking, based on random K-fold validation, is established for COVID-19 mortality risk forecasting and assessment, involving Machine Learning (ML) models, and the adopted Classical and Bayesian semilinear estimation approach. This empirical analysis also determines the ML models favored by the spatial multivariate Functional Data Analysis (FDA) framework. The results could be extrapolated to other countries.

Evaluating artificial shelter arrays as a minimally invasive monitoring tool for the hellbender Cryptobranchus alleganiensis

Hellbenders Cryptobranchus alleganiensis are critically imperiled amphibians throughout the eastern USA. Rock-lifting is widely used to monitor hellbenders but can severely disturb habitat. We asked whether artificial shelter occupancy (the proportion of occupied shelters in an array) would function as a proxy for hellbender abundance and thereby serve as a viable alternative to rock-lifting. We hypothesized that shelter occupancy would vary spatially in response to hellbender density, natural shelter density, or both, and would vary temporally with hellbender seasonal activity patterns and time since shelter deployment. We established shelter arrays (n = 30 shelters each) in 6 stream reaches and monitored them monthly for up to 2 yr. We used Bayesian mixed logistic regression and model ranking criteria to assess support for hypotheses concerning drivers of shelter occupancy. In all reaches, shelter occupancy was highest from June-August each year and was higher in Year 2 relative to Year 1. Our best-supported model indicated that the extent of boulder and bedrock (hereafter, natural shelter) in a reach mediated the relationship between hellbender abundance and shelter occupancy. More explicitly, shelter occupancy was positively correlated with abundance when natural shelter covered <20% of a reach, but uncorrelated with abundance when natural shelter was more abundant. While shelter occupancy should not be used to infer variation in hellbender relative abundance when substrate composition varies among reaches, we showed that artificial shelters can function as valuable monitoring tools when reaches meet certain criteria, though regular shelter maintenance is critical.

SPARQL QUERY GENERATION FOR COMPLEX QUESTION ANSWERING WITH BERT AND BILSTM-BASED MODEL

In this paper we describe question answering system for answering of complex questions over Wikidata knowledge base. Unlike simple questions, which require extraction of single fact from the knowledge base, complex questions are based on more than one triplet and need logical or comparative reasoning. The proposed question answering system translates a natural language question into a query in SPARQL language, execution of which gives an answer. The system includes the models which define the SPARQL query template corresponding to the question and then fill the slots in the template with entities, relations and numerical values. For entity detection we use BERTbased sequence labelling model. Ranking of candidate relations is performed in two steps with BiLSTM and BERT-based models. The proposed models are the first solution for LC-QUAD2.0 dataset. The system is capable of answering complex questions which involve comparative or boolean reasoning.

Digital Space of Leading Universities: The Comparative Analysis of Sites

The article presents the results of a comparative and typological analysis of open data websites of the leading Russian and foreign universities. The analysis has been carried out in order to assess the state of the digital environment of universities in the areas of career guidance of students, the development of students’ competencies and career support of graduates. The authors adduce an empirical evidence of the key problems of digitalization of Russian universities such as the lack of interactive innovative forms of online support of applicants by Russian universities; the lack of digital footprints of the educational process, especially video lessons; depersonalization of digital support of training; non-competitive descriptions of elective courses, the lack of educational case studies as a main technology for the implementation of practice-oriented tasks; the underdevelopment of interaction as a basic principle for providing online support of the educational process. As the prospects of the study, the article makes the case for constructing a model ranking universities according to the level of digital space development.

Genetic evaluation of clinical mastitis traits in Holstein cattle

The results obtained from different models for predicting breeding values for clinical mastitis (CM) in Holstein cattle were compared. CM was recorded in 30 882 lactations of 12 793 cows in 8 herds from 1996 to 2016. CM was considered either as an all-or-none binary trait (0 – absence of mastitis; 1 – at least one case of CM per lactation) or as the number of cases. CM is recorded in the first 150 days of lactation or throughout the entire course of lactation. Breeding values were predicted with a single-trait repeatability model and with a bivariate animal model, where CM during the 1^st lactation and CM during the 2^nd and later lactations were considered as two different traits. Estimated heritability ranged from 0.06 for CM as a 0/1 trait during the first 150 days of lactation in the repeatability model to 0.12 for the number of CM cases during the 1^st lactation in the bivariate model. Ranking of the sires with 15 or more daughters and of the cows was performed according to their breeding values, and Spearman correlation coefficients were calculated. Rank correlations of breeding values from the repeatability and bivariate models were stronger for the same parts of lactation (150 vs 305 days; 0.88–0.96) and for the repeatability model and the bivariate model for the 2^nd and later lactations (0.95–0.98). Trends of average male and female breeding values according to their birth year were used to assess genetic changes in the population. The average breeding values declined slowly for cows born since 2003 but stayed above neutral value, thus indicating permanent genetic deterioration of mastitis resistance.

NEPRE: a Scoring Function for Protein Structures based on Neighbourhood Preference

ABSTRACTProtein structure prediction relies on two major components, a method to generate good models that are close to the native structure and a scoring function that can select the good models. Based on the statistics from known structures in the protein data bank, a statistical energy function is derived to reflect the amino acid neighbourhood preferences. The neighbourhood of one amino acid is defined by its contacting residues, and the energy function is determined by the neighbhoring residue types and relative positions. A scoring algorithm, Nepre, has been implemented and its performance was tested with several decoy sets. The results show that the Nepre program can be applied in model ranking to improve the success rate in structure predictions.

Bayesian performance evaluation of evapotranspiration models for an arid region in northwestern China

Evapotranspiration (ET) is a major component of the land surface process involved in energy fluxes and balance, especially in the hydrological cycle of agricultural ecosystems. While many models have been developed to estimate ET, there has been no agreement on which model has the best performance. In this study, we evaluate four widely used ET models (i.e., the Shuttleworth Wallace (SW) model, Penman-Monteith (PM) model, Priestley-Taylor and Flint-Childs (PT-FC) model, and Advection-Aridity (AA) model) by using half-hourly ET observations obtained at a spring maize field in an arid region. The model evaluation is based on Bayesian model comparison and ranking using the Bayesian model evidence (BME), which balances between goodness-of-fit to data and model complexity. The BME-based model ranking (from the best to the worst) is SW, PM, PT-FC, and AA. The residuals between observations and corresponding model simulations are also analyzed, and the same model ranking is also obstained by using residual-based statistics, i.e., the coefficient of determination (R2), index of agreement (IA), root mean square error (RMSE) and model efficiency (EF). The PM and SW models overestimate ET, whereas the PT-FC and AA models underestimate ET in the study period. The four models also underestimate ET during the periods of partial crop cover. Especially during the late maturity stage, the PT-FC and AA models consistently produce an underestimation, and provide the worst simulated ET. As a result, at the half-hourly time scale, the SW model is the best model and recommend as the first choice for evaluating ET of spring maize in arid desert oasis areas.