Likelihood Metric Research Articles

Taxonomy of Tobrilidae species from the Alkaline Lakes of the western Nebraska Sandhills.

Six distinct COI mitochondrial Haplotype Groups (HG) are morphologically, ecologically, and genetically characterized from the aquatic nematode family Tobrilidae. Collection locations included the extreme habitats of the Alkaline Lakes in the western Nebraska Sandhills and the contaminated stream, Johnson Creek, bordering the AltEn 2021 catastrophic pesticide release near the village of Mead in eastern Nebraska. Maximum likelihood and genetic distance metrics supported the genetic integrity of the haplotype groups. Discriminant function analysis of COI haplotype group datasets of combined morphological characters and soil chemistry attributes for both male and female Tobrilidae were classified correctly in all but one case. Scanning electron microscopy revealed new details about amphid apertures, male supplements, and spicules. Partial 18S gene phylogeny suggests that the genus Semitobrilus may not be a member of the subfamily Neotobrilinae, and three specimens in the 226 tobrilid dataset provide evidence of incongruence between COI and 18S derived phylogenies. Given the strong signal provided by the environmental chemistry data, tobrilid mitochondrial haplotypes may well have value as environmental indicators.

Using probabilistic machine learning to better model temporal patterns in parameterizations: a case study with the Lorenz 96 model

Abstract. The modelling of small-scale processes is a major source of error in weather and climate models, hindering the accuracy of low-cost models which must approximate such processes through parameterization. Red noise is essential to many operational parameterization schemes, helping model temporal correlations. We show how to build on the successes of red noise by combining the known benefits of stochasticity with machine learning. This is done using a recurrent neural network within a probabilistic framework (L96-RNN). Our model is competitive and often superior to both a bespoke baseline and an existing probabilistic machine learning approach (GAN, generative adversarial network) when applied to the Lorenz 96 atmospheric simulation. This is due to its superior ability to model temporal patterns compared to standard first-order autoregressive schemes. It also generalizes to unseen scenarios. We evaluate it across a number of metrics from the literature and also discuss the benefits of using the probabilistic metric of hold-out likelihood.

A framework for designing vulnerability metrics

Vulnerability analysis has long been used to evaluate the security posture of a system. Different approaches, including vulnerability graphs and various vulnerability metrics, have been used to study the vulnerability landscape and provide security analysts with cyber situational awareness. However, most current solutions still lack a principled approach to quantifying various dimensions of known vulnerabilities in a way that can easily adapt to different applicative domains and operating conditions. To address this limitation, we introduce a vulnerability metrics framework that extends and generalizes our previous metrics for evaluating the exploitation likelihood of a vulnerability and the exposure factor of system components to vulnerability exploits. We argue that the factors influencing these metrics and their relative weights depend on the specific applicative domain, defender’s priorities, and attacker’s knowledge. Thus, instead of providing a static set of equations, we establish a framework for instantiating the equations that best model the scenario being considered. We combine likelihood and exposure factor metrics into a severity score that allows us to rank vulnerabilities. In our evaluation, we demonstrate that ranking vulnerabilities solely based on their CVSS scores is not sufficient for effective prioritization, due to the limited number of possible distinct severity values compared to the sheer number of existing vulnerabilities. We define a ranking quality score and show that considering additional information about vulnerabilities helps refine their ranking, providing more actionable intelligence to security analysts.

Radiometric Identification of Signals by Matched Whitening Transform

Radiometric identification is the problem of attributing a signal to a specific source. In this work, a radiometric identification algorithm is developed using the whitening transformation. The approach stands out from the more established methods in that it works directly on the raw IQ data and hence is featureless. As such, the commonly used dimensionality reduction algorithms do not apply. The premise of the idea is that a data set is “most white” when projected on its own whitening matrix than on any other. In practice, transformed data are never strictly white since the training and the test data differ. The Förstner-Moonen measure that quantifies the similarity of covariance matrices is used to establish the degree of whiteness. The whitening transform that produces a data set with the minimum Förstner-Moonen distance to a white noise process is the source signal. The source is determined by the output of the mode function operated on the Majority Vote Classifier decisions. Using the Förstner-Moonen measure presents a different perspective compared to maximum likelihood and Euclidean distance metrics. The whitening transform is also contrasted with the more recent deep learning approaches that are still dependent on feature vectors with large dimensions and lengthy training phases. It is shown that the proposed method is simpler to implement, requires no features vectors, needs minimal training and because of its non-iterative structure is faster than existing approaches.

Measuring the likelihood of VR visual fatigue through ocular biomechanics

Immersion in virtual reality is still linked to symptoms of visual fatigue such as eye strain, dizziness, and overall discomfort. Studies have investigated visual fatigue through pre- and post-immersion tests of the visual function. In this work, we extend on our previous study and derive a visual fatigue likelihood metric using biomechanical analysis. Previously, we have investigated the effect of VR on the vergence system during immersion. The proposed visual fatigue metric exhibited a significant correlation to vergence angle variability which was previously linked to vergence accommodation conflict in VR. We also discuss subjective feedback and its relationship with the proposed visual fatigue metric.

The Population Pharmacokinetics of Meropenem in Adult Patients With Rifampicin-Sensitive Pulmonary Tuberculosis.

Background: Meropenem is being investigated for repurposing as an anti-tuberculosis drug. This study aimed to develop a meropenem population pharmacokinetics model in patients with pulmonary tuberculosis and identify covariates explaining inter-individual variability. Methods: Patients were randomized to one of four treatment groups: meropenem 2 g three times daily plus oral rifampicin 20 mg/kg once daily, meropenem 2 g three times daily, meropenem 1 g three times daily, and meropenem 3 g once daily. Meropenem was administered by intravenous infusion over 0.5–1 h. All patients also received oral amoxicillin/clavulanate together with each meropenem dose, and treatments continued daily for 14 days. Intensive plasma pharmacokinetics sampling over 8 h was conducted on the 14th day of the study. Nonlinear mixed-effects modeling was used for data analysis. The best model was chosen based on likelihood metrics, goodness-of-fit plots, and parsimony. Covariates were tested stepwise. Results: A total of 404 concentration measurements from 49 patients were included in the analysis. A two-compartment model parameterized with clearance (CL), inter-compartmental clearance (Q), and central (V1) and peripheral (V2) volumes of distribution fitted the data well. Typical values of CL, Q, V1, and V2 were 11.8 L/h, 3.26 L/h, 14.2 L, and 3.12 L, respectively. The relative standard errors of the parameter estimates ranged from 3.8 to 35.4%. The covariate relations included in the final model were creatinine clearance on CL and allometric scaling with body weight on all disposition parameters. An effect of age on CL as previously reported could not be identified. Conclusion: A two-compartment model described meropenem population pharmacokinetics in patients with pulmonary tuberculosis well. Covariates found to improve model fit were creatinine clearance and body weight but not rifampicin treatment. The final model will be used for an integrated pharmacokinetics/pharmacodynamics analysis linking meropenem exposure to early bactericidal activity.

Принципы построения программно-аппаратного комплекса для интеллектуальной поддержки экипажа и оценки уровня его подготовки

Presented is a new approach to aircraft crew intelligent support, which is based on comparing flight fragments (maneuvers) under study with the relevant patterns contained in the database and representing the system “empirical intelligence”. Principal components of this approach are four new metrics for comparing flight fragments, viz.: the Euclidean metric in the space of wavelet coefficients; the likelihood metric of eigenvalue trajectories for transformations of activity parameters; the Kohonen metric in the space of wavelet coefficients; the likelihood metric for comparing gaze trajectories. Features of the presented approach are: the presence of an “intelligent component” that is contained in empirical data and can be flexibly changed as they accumulate; the use of integral comparisons of the flight fragments under study and video oculography data with relevant patterns of various types and performance quality from a specialized database, with transferring characteristics of the nearest pattern from this specialized database to the fragment under study; applying a complex combination of the methods for stochastic processes analysis and multivariate statistical techniques.

Defining and evaluating predictions of joint species distribution models

Abstract Joint species distribution models (JSDMs) simultaneously model the distributions of multiple species, while accounting for residual co‐occurrence patterns. Despite increasing adoption of JSDMs in the literature, the question of how to define and evaluate JSDM predictions has only begun to be explored. We define four different JSDM prediction types that correspond to different aspects of species distribution and community assemblage processes. Marginal predictions are environment‐only predictions akin to predictions from single‐species models; joint predictions simultaneously predict entire community assemblages; and conditional marginal and conditional joint predictions are made at the species or assemblage level, conditional on the known occurrence state of one or more species at a site. We define five different classes of metrics that can be used to evaluate these types of predictions: threshold‐dependent, threshold‐independent, community dissimilarity, species richness and likelihood metrics. We illustrate different prediction types and evaluation metrics using a case study in which we fit a JSDM to a frog occurrence dataset collected in Melbourne, Australia. Joint species distribution models present opportunities to investigate the facets of species distribution and community assemblage processes that are not possible to explore with single‐species models. We show that there are a variety of different metrics available to evaluate JSDM predictions, and that choice of prediction type and evaluation metric should closely match the questions being investigated.

An Agent-Based Model for No Fault Found Events

This paper develops an agent-based model to estimate the likelihood of the No Fault Found (NFF) phenomenon. NFF events are critical problems as the root cause of an issue cannot be diagnosed. This makes it difficult to quantify associated costs as potentially serviceable equipment is being returned for repair. Here, the authors first map the dynamic interactions experienced during NFF events across a predefined supply chain. The motivation is to understand component flow by evaluating the root causes of NFF in a more proactive way across the supply network. The data is collected using semi-structured interviews with 12 maintenance-related experts to understand how NFF manifests during maintenance across 7 UK organisations. Then, the NFF phenomenon across the supply chain is reflected using an agent-based model; to support decision-making at the design level, on component selection in terms of product reliability, and at the service level to support decision-making. Monitoring the NFF likelihood metric can help manage the problem, whilst demonstrate how qualitative information can be used to achieve maintenance objectives.

Modified ordered successive interference cancellation MIMO detection using low complexity constellation search

High spectral efficiency, large capacity, and high data rate make large-scale Multiple-Input Multiple-Output (MIMO) systems more prominent in 5G communication. Existing MIMO detection schemes like Maximum Likelihood (ML) achieves optimal performance, but it is significantly affected by exponentially increasing computational complexity. On the other hand, linear MIMO detection techniques have low complexity. However, these cannot be implemented in practical systems due to poor performance. Ordered Successive Interference Cancellation (OSIC) addresses these issues as it possesses intermediate and suboptimal performance but it faces erroneous symbols due to wrong selection of order. At first, OSIC K-correction algorithm is employed to resolve OSIC issues by replacing the user-defined K symbols in the search subspace of OSIC symbol vector and detects the symbol in terms of better likelihood metric. In addition, it proposes Reduced Complexity (RC) OSIC K-correction scheme that further reduces the computational complexity in OSIC K-correction by exploiting the deviation vector which focuses to opt and detect only the most erroneous symbols. The proposed algorithms offer an adequate trade-off between computational complexity and detection performance.

Conserving native trees increases native bird diversity and community composition on commercial office developments

Abstract In cities, woody vegetation provides critical shelter, nesting and foraging habitat for bird species of interest. Human actions—including development and landscaping choices—determine vegetation community composition and structure, making these choices critically important to urban bird conservation. A better understanding of how bird communities are impacted by parcel-scale actions can help guide policy and management best practices to improve matrix habitat quality and quantity. Here, I examined how bird habitat use varies along a vegetation gradient created by different development and landscaping choices. I surveyed 20 commercial office developments near Seattle in the Puget Trough region of Washington, USA selected using stratified random sampling, where I quantified bird communities and observed feeding behavior. I used GLMM and PERMANOVA models with data likelihood metrics to identify the best supported variables for bird site use, along with TITAN models to identify changes in community composition along environmental gradients. I found that measures of bird effective species richness and bird community are positively influenced by the presence of more native conifers, including the presence of a stand predating development and the height and density of native conifers. Measures of the native bird community are negatively influenced by higher non-native tree density. In contrast to prior research, top-down landscape-scale variables did not explain variation in measures of the bird community on office developments. Importantly, I found that birds are associated with the same habitat on office developments as observed elsewhere. Together, my findings suggest an important role for developers, land owners, landscape architects, and tree protection policy in bird conservation.

Optimising forecasting models for inventory planning

Inaccurate forecasts can be costly for company operations, in terms of stock-outs and lost sales, or over-stocking, while not meeting service level targets. The forecasting literature, often disjoint from the needs of the forecast users, has focused on providing optimal models in terms of likelihood and various accuracy metrics. However, there is evidence that this does not always lead to better inventory performance, as often the translation between forecast errors and inventory results is not linear. In this study, we consider an approach to parametrising forecasting models by directly considering appropriate inventory metrics and the current inventory policy. We propose a way to combine the competing multiple inventory objectives, i.e. meeting demand, while eliminating excessive stock, and use the resulting cost function to identify inventory optimal parameters for forecasting models. We evaluate the proposed parametrisation against established alternatives and demonstrate its performance on real data. Furthermore, we explore the connection between forecast accuracy and inventory performance and discuss the extent to which the former is an appropriate proxy of the latter.

HyDRA-H: Simultaneous Hybrid Retrieval of Exoplanetary Emission Spectra

Abstract High-resolution Doppler spectroscopy has been used to detect several chemical species in exoplanetary atmospheres. Such detections have traditionally relied on cross correlation of observed spectra against spectral model templates, an approach that is successful for detecting chemical species but not optimized for constraining abundances. Recent work has explored ways to perform atmospheric retrievals on high-resolution spectra (HRS) and combine them with retrievals routinely performed for low-resolution spectra (LRS) by developing a mapping from the cross-correlation function to a likelihood metric. We build upon previous studies and report HyDRA-H, a hybrid retrieval code for simultaneous analysis of low- and high- resolution thermal emission spectra of exoplanets in a fully Bayesian approach. We demonstrate HyDRA-H on the hot Jupiter HD 209458b as a case study. We validate our HRS retrieval capability by confirming previous results and report a simultaneous hybrid retrieval using both HRS and LRS data. The LRS data span the HST WFC3 (1.1–1.7 μm) and Spitzer photometry (IRAC 3.6–8 μm) bands, while the HRS data were obtained with CRIRES on VLT at 2.3 μm. The constraints on the composition and temperature profiles for the hybrid retrieval are more stringent than retrievals with either LRS or HRS data sets individually. We retrieve abundances of and , and C/O = , consistent with previous works. We constrain the photospheric temperature to be K, consistent with the equilibrium temperature. Our results demonstrate the significant advantages of hybrid retrievals by combining strengths of both HRS and LRS observations, which probe complementary aspects of exoplanetary atmospheres.

Meteorological Connectivity from Regions of High Biodiversity within the McMurdo Dry Valleys of Antarctica

AbstractMeteorological connectivity between biological hot spots of the McMurdo Dry Valleys (MDVs) of Antarctica is thought to play a role in species distribution and abundance through the aeolian transport of bioaerosols. Understanding the potential role of such meteorological connectivity requires an understanding of near-surface wind flow within and between valley airsheds. To address this, we applied Lagrangian wind trajectory modeling to mesoscale (spatial resolution of ~1 km) weather model output to predict connectivity pathways, focusing on regions of high biodiversity. Our models produce maps of a likelihood metric of wind connectivity that demonstrate the synoptic and mesoscale dependence of connections between local, near-local, and nonlocal areas on wind transport, modulated by synoptic weather and topographic forcing. These connectivity areas can have spatial trends modulated by the synoptic weather patterns and locally induced topographically forced winds. This method is transferrable to other regions of Antarctica for broader terrestrial, coastal, and offshore ecological connectivity research. Also, our analysis and methods can inform better placement of aeolian dust and bioaerosol samplers in the McMurdo Dry Valleys, provide preliminary guidelines behind the meteorological controls of sediment transport and smaller particle distribution, and present quantifiable knowledge informing new hypotheses around the potential of wind acting as a physical driver for biological connectivity in the MDVs.

Multi-outcome longitudinal small area estimation – a case study

ABSTRACTA recent paper [Thibaudeau, Slud, and Gottschalck (2017). Modeling log-linear conditional probabilities for estimation in surveys. The Annals of Applied Statistics, 11, 680–697] proposed a ‘hybrid’ method of survey estimation combining coarsely cross-classified design-based survey-weighted totals in a population with loglinear or generalised-linear model-based conditional probabilities for cells in a finer cross-classification. The models were compared in weighted and unweighted forms on data from the US Survey of Income and Program Participation (SIPP), a large national longitudinal survey. The hybrid method was elaborated in a book-chapter [Thibaudeau, Slud, & Cheng (2019). Small-area estimation of cross-classified gross flows using longitudinal survey data. In P. Lynn (Ed.), Methodology of longitudinal surveys II. Wiley] about estimating gross flows in (two-period) longitudinal surveys, by considering fixed versus mixed effect versions of the conditional-probability models and allowing for 3 or more outcomes in the later-period categories used to define gross flows within generalised logistic regression models. The methodology provided for point and interval small-area estimation, specifically area-level two-period labour-status gross-flow estimation, illustrated on a US Current Population Survey (CPS) dataset of survey respondents in two successive months in 16 states. In the current paper, that data analysis is expanded in two ways: (i) by analysing the CPS dataset in greater detail, incorporating multiple random effects (slopes as well as intercepts), using predictive as well as likelihood metrics for model quality, and (ii) by showing how Bayesian computation (MCMC) provides insights concerning fixed- versus mixed-effect model predictions. The findings from fixed-effect analyses with state effects, from corresponding models with state random effects, and fom Bayes analysis of posteriors for the fixed state-effects with other model coefficients fixed, all confirm each other and support a model with normal random state effects, independent across states.

Estimating parameters of nonlinear dynamic systems in pharmacology using chaos synchronization and grid search

Bridging fundamental approaches to model optimization for pharmacometricians, systems pharmacologists and statisticians is a critical issue. These fields rely primarily on Maximum Likelihood and Extended Least Squares metrics with iterative estimation of parameters. Our research combines adaptive chaos synchronization and grid search to estimate physiological and pharmacological systems with emergent properties by exploring deterministic methods that are more appropriate and have potentially superior performance than classical numerical approaches, which minimize the sum of squares or maximize the likelihood. We illustrate these issues with an established model of cortisol in human with nonlinear dynamics. The model describes cortisol kinetics over time, including its chaotic oscillations, by a delay differential equation. We demonstrate that chaos synchronization helps to avoid the tendency of the gradient-based optimization algorithms to end up in a local minimum. The subsequent analysis illustrates that the hybrid adaptive chaos synchronization for estimation of linear parameters with coarse-to-fine grid search for optimal values of non-linear parameters can be applied iteratively to accurately estimate parameters and effectively track trajectories for a wide class of noisy chaotic systems.

Новые подходы к оцениванию работы операторов сложных технических систем

Presented are new approaches for supporting the outcome grading for activities of operators of complex technical systems, which are based on comparisons of current exercises with the activity database patterns in both the wavelet representation metric associated with time series of activity parameters and the likelihood metric of eigenvalue trajectories for these parameters transforms as well as on probabilistic assessments of skill class recognition using sample distribution functions of exercise distances to cluster centers in a scaling space and Bayesian likelihood estimations with the aid of probabilistic profile of staying in activity parameter ranges. These techniques have demonstrated the capabilities of recognizing sets of abnormal exercises and detection of parameters characterizing operator mistakes to reveal the causes of abnormality. The techniques in question overcome limitations of existing methods and provide advantages over manual data analysis since they greatly reduce the combinatorial enumeration of the options considered.

Optimising Forecasting Models for Inventory Planning

Inaccurate forecasts can be costly for company operations, in terms of stock-outs and lost sales, or over-stocking, while not meeting service level targets. The forecasting literature, often disjoint from the needs of the forecast users, has focused on providing optimal models in terms of likelihood and various accuracy metrics. However, there is evidence that this does not always lead to better inventory performance, as often the translation between forecast errors and inventory results is not linear. In this study, we consider an approach to parametrising forecasting models by directly considering appropriate inventory metrics and the current inventory policy. We propose a way to combine the competing multiple inventory objectives, i.e. meeting demand, while eliminating excessive stock, and use the resulting cost function to identify inventory optimal parameters for forecasting models. We evaluate the proposed parametrisation against established alternatives and demonstrate its performance on real data. Furthermore, we explore the connection between forecast accuracy and inventory performance and discuss the extent to which the former is an appropriate proxy of the latter.

Combining probability with qualitative degree-of-certainty metrics in assessment

Reports of the Intergovernmental Panel on Climate Change (IPCC) employ an evolving framework of calibrated language for assessing and communicating degrees of certainty in findings. A persistent challenge for this framework has been ambiguity in the relationship between multiple degree-of-certainty metrics. We aim to clarify the relationship between the likelihood and confidence metrics used in the Fifth Assessment Report (2013), with benefits for mathematical consistency among multiple findings and for usability in downstream modeling and decision analysis. We discuss how our proposal meshes with current and proposed practice in IPCC uncertainty assessment.

Potential geography and productivity of “Hass” avocado crops in Colombia estimated by ecological niche modeling

Abstract Variation in environmental suitability across the geographic range of a species is common, with consequences for population health and production. Ecological niche modeling (ENM) is an effort to estimate environmental requirements of species based on associations between known geographic occurrences and environmental conditions at those sites, to allow estimation of the potential distribution of the species. We developed an ENM for Hass avocado production fields in Colombia, in which we related the distribution of avocado production lots to environmental information derived from multitemporal MODIS satellite imagery and digital elevation models. Models were calibrated using maximum entropy approaches, using likelihood and information content metrics to optimize model selection; model performance was measured in terms of omission error with respect to independent testing data. Two models were selected that presented best combinations of high information content, significance, and performance based on lower values of Akaike Information Criterion (AICc), partial receiver operating characteristic (partialROC), and low omission of independent testing data records. Suitability values from those models were compared to Hass avocado production; models were able to identify both novel areas with potential for cultivation, and sites where this crop has been planted in relatively unsuitable areas. A statistically significant positive relationship existed between suitability values from the models and avocado production, with different relationships for three technological levels. ENM approaches thus offer a useful and novel tool for exploring and designing sustainable agriculture systems, identifying areas with environmental potential in Colombia and currently not planted with Hass avocado.