Commentary to: a cross-validation-based approach for delimiting reliable home range estimates.
Ontology highlight
ABSTRACT:
SUBMITTER: Dougherty ER
PROVIDER: S-EPMC6042211 | biostudies-other | 2018
REPOSITORIES: biostudies-other
Publications
Commentary to: a cross-validation-based approach for delimiting reliable home range estimates.
Dougherty Eric R ER de Valpine Perry P Carlson Colin J CJ Blackburn Jason K JK Getz Wayne M WM
Movement ecology 20180712
<h4>Background</h4>Continued exploration of the performance of the recently proposed cross-validation-based approach for delimiting home ranges using the Time Local Convex Hull (T-LoCoH) method has revealed a number of issues with the original formulation.<h4>Main text</h4>Here we replace the ad hoc cross-validation score with a new formulation based on the total log probability of out-of-sample predictions. To obtain these probabilities, we interpret the normalized LoCoH hulls as a probability ...[more]
Similar Datasets
Project description:Variation in spatial and temporal distribution of resources drives animal movement patterns. Links between ecology and behavior are particularly salient for the multilevel society of hamadryas baboons, in which social units cleave and coalesce over time in response to ecological factors. Here, we used data from GPS collars to estimate home range size and assess temporal patterns of sleeping site use in a band of hamadryas baboons in Awash National Park, Ethiopia. We used GPS data derived from 2 to 3 collared baboons over three 8-12-month collaring intervals to estimate annual and monthly home ranges using kernel density estimators (KDEs) and minimum convex polygons (MCPs). The 95% KDE home range was 64.11 km2 for Collaring Interval I (July 2015-March 2016), 85.52 km2 for Collaring Interval II (October 2016-October 2017), 76.43 km2 for Collaring Interval III (July 2018-May 2019), and 75.25 km2 across all three collaring intervals. MCP home ranges were 103.46 km2 for Collaring Interval I, 97.90 km2 for Collaring Interval II, 105.22 km2 for Collaring Interval III, and 129.33 km2 overall. Ninety-five percent KDE home range sizes did not differ across months, nor correlate with temperature or precipitation, but monthly MCP home ranges increased with monthly precipitation. Our data also revealed a southward home range shift over time and seven previously unknown sleeping sites, three of which were used more often during the wet season. Band cohesion was highest during dry months and lowest during wet months, with fissioning occurring more frequently at higher temperatures. One pair of collared individuals from Collaring Interval III spent 95% of nights together, suggesting they were members of the same clan. Our results both suggest that previous studies have underestimated the home range size of hamadryas baboons and highlight the benefits of remote data collection.
| S-EPMC8244037 | biostudies-literature
Project description:The rate of self-fertilization (that is, selfing) is a key evolutionary parameter in hermaphroditic species, yet obtaining accurate estimates of selfing rates in natural populations can be technically challenging. Most published estimates are derived from population-level heterozygote deficiency (that is, FIS) or identity disequilibria (for example, the software RMES (robust multilocus estimate of selfing)). These indirect methods can be applied to population genetic survey data, whereas direct methods using progeny arrays require much larger data sets that are often difficult to collect in natural populations or even require captive breeding. Unfortunately, indirect methods rely on assumptions that can be problematic, such as negating biparental inbreeding, inbreeding disequilibrium and (for FIS) the presence of null alleles. The performance of indirect estimates against progeny-array estimates is still largely unknown. Here we used both direct progeny-array and indirect population-level methods to estimate the selfing rate in a single natural population of the simultaneously hermaphroditic freshwater snail Radix balthica throughout its reproductive lifespan using 10 highly polymorphic microsatellites. We found that even though progeny arrays (n=1034 field-collected embryos from 60 families) did not reveal a single selfed embryo, FIS-based selfing rates (n=316 adults) were significantly positive in all 6 sequential population samples. Including a locus with a high frequency of null alleles further biased FIS-based estimates. Conversely, RMES-based estimates were very similar to progeny-array estimates and proved insensitive to null alleles. The assumptions made by RMES were thus either met or irrelevant in this particular population, making RMES a valid, cost-efficient alternative to progeny arrays.
| S-EPMC5436024 | biostudies-literature
Project description:Purpose ?To determine the reliability of wrist range of motion (WROM) measurements based on digital photographs taken by patients at home compared with traditional measurements done in the office with a goniometer. Methods ?Sixty-nine postoperative patients were enrolled in this study at least 3 months postoperatively. Active and passive wrist flexion/extension and radial/ulnar deviation were recorded by one of the two attending surgeons with a 1-degree resolution goniometer at the last postoperative office visit. Patients were provided an illustrated instruction sheet detailing how to take digital photographic images at home in six wrist positions (active and passive flexion/extension, and radial/ulnar deviation). Wrist position was measured from digital images by both the attending surgeons in a randomized, blinded fashion on two separate occasions greater than 2 weeks apart using the same goniometer. Reliability analysis was performed using the intraclass correlation coefficient to assess agreement between clinical and photography-based goniometry, as well as intra- and interobserver agreement. Results ?Out of 69 enrolled patients, 30 (43%) patients sent digital images. Of the 180 digital photographs, only 9 (5%) were missing or deemed inadequate for WROM measurements. Agreement between clinical and photography-based measurements was "almost perfect" for passive wrist flexion/extension and "substantial" for active wrist flexion/extension and radial/ulnar deviation. Inter- and intraobserver agreement for the attending surgeons was "almost perfect" for all measurements. Discussion ?This study validates a photography-based goniometry protocol allowing accurate and reliable WROM measurements without direct physician contact. Passive WROM was more accurately measured from photographs than active WROM. This study builds on previous photography-based goniometry literature by validating a protocol in which patients or their families take and submit their own photographs. Clinical Relevance ?Patient-performed photography-based goniometry represents an alternative to traditional clinical goniometry that could enable longer-term follow-up, overcome travel-related impediments to office visits, improve convenience, and reduce costs for patients.
| S-EPMC5658219 | biostudies-literature
Project description:BackgroundGenerally, QSAR modelling requires both model selection and validation since there is no a priori knowledge about the optimal QSAR model. Prediction errors (PE) are frequently used to select and to assess the models under study. Reliable estimation of prediction errors is challenging - especially under model uncertainty - and requires independent test objects. These test objects must not be involved in model building nor in model selection. Double cross-validation, sometimes also termed nested cross-validation, offers an attractive possibility to generate test data and to select QSAR models since it uses the data very efficiently. Nevertheless, there is a controversy in the literature with respect to the reliability of double cross-validation under model uncertainty. Moreover, systematic studies investigating the adequate parameterization of double cross-validation are still missing. Here, the cross-validation design in the inner loop and the influence of the test set size in the outer loop is systematically studied for regression models in combination with variable selection.MethodsSimulated and real data are analysed with double cross-validation to identify important factors for the resulting model quality. For the simulated data, a bias-variance decomposition is provided.ResultsThe prediction errors of QSAR/QSPR regression models in combination with variable selection depend to a large degree on the parameterization of double cross-validation. While the parameters for the inner loop of double cross-validation mainly influence bias and variance of the resulting models, the parameters for the outer loop mainly influence the variability of the resulting prediction error estimate.ConclusionsDouble cross-validation reliably and unbiasedly estimates prediction errors under model uncertainty for regression models. As compared to a single test set, double cross-validation provided a more realistic picture of model quality and should be preferred over a single test set.
| S-EPMC4260165 | biostudies-literature
Project description:Background: There is a pressing need for standardized measures to assess the quality of home-based serious illness care. Currently, there are no validated quality measures that are specific to home-based serious illness programs (SIPs) and the unique needs of their patients. Objective: To develop and evaluate standardized survey-based measures of serious illness care experiences for assessing and comparing quality of home-based serious illness care programs. Methods: From October 2019 through January 2020, we administered a survey to patients who received care from 32 home-based SIPs across the United States. Using the 2263 survey responses, we assessed item performance and constructed composite measures via factor analysis, evaluated item-scale correlations, estimated reliability, and examined validity by regressing overall ratings and willingness to recommend care on each composite. Results: The overall survey response rate was 36%. Confirmatory factor analyses supported five composite quality measures: Communication, Care Coordination, Help for Symptoms, Planning for Care, and Support for Family and Friends. Cronbach's alpha estimates for the composite measures ranged from 0.69 to 0.85, indicating adequate internal consistency in assessing their underlying constructs. Interprogram reliability ranged from 0.67 to 0.80 at 100 completed surveys per measure, meeting common standards for distinguishing between programs' performance. Together, the composites explained 45% of the variance in patients' overall care ratings. Communication, Care Coordination, and Planning for Care were the strongest predictors of overall ratings. Conclusion: Our analyses provide evidence of the feasibility, reliability, and validity of proposed survey-based measures to assess the quality of home-based serious illness care from the perspective of patients and their families.
| S-EPMC9145570 | biostudies-literature
Project description:BackgroundWith the large amount of biological data that is currently publicly available, many investigators combine multiple data sets to increase the sample size and potentially also the power of their analyses. However, technical differences ("batch effects") as well as differences in sample composition between the data sets may significantly affect the ability to draw generalizable conclusions from such studies.FocusThe current study focuses on the construction of classifiers, and the use of cross-validation to estimate their performance. In particular, we investigate the impact of batch effects and differences in sample composition between batches on the accuracy of the classification performance estimate obtained via cross-validation. The focus on estimation bias is a main difference compared to previous studies, which have mostly focused on the predictive performance and how it relates to the presence of batch effects.DataWe work on simulated data sets. To have realistic intensity distributions, we use real gene expression data as the basis for our simulation. Random samples from this expression matrix are selected and assigned to group 1 (e.g., 'control') or group 2 (e.g., 'treated'). We introduce batch effects and select some features to be differentially expressed between the two groups. We consider several scenarios for our study, most importantly different levels of confounding between groups and batch effects.MethodsWe focus on well-known classifiers: logistic regression, Support Vector Machines (SVM), k-nearest neighbors (kNN) and Random Forests (RF). Feature selection is performed with the Wilcoxon test or the lasso. Parameter tuning and feature selection, as well as the estimation of the prediction performance of each classifier, is performed within a nested cross-validation scheme. The estimated classification performance is then compared to what is obtained when applying the classifier to independent data.
| S-EPMC4072626 | biostudies-literature
Project description:Online estimators update a current estimate with a new incoming batch of data without having to revisit past data thereby providing streaming estimates that are scalable to big data. We develop flexible, ensemble-based online estimators of an infinite-dimensional target parameter, such as a regression function, in the setting where data are generated sequentially by a common conditional data distribution given summary measures of the past. This setting encompasses a wide range of time-series models and, as special case, models for independent and identically distributed data. Our estimator considers a large library of candidate online estimators and uses online cross-validation to identify the algorithm with the best performance. We show that by basing estimates on the cross-validation-selected algorithm, we are asymptotically guaranteed to perform as well as the true, unknown best-performing algorithm. We provide extensions of this approach including online estimation of the optimal ensemble of candidate online estimators. We illustrate excellent performance of our methods using simulations and a real data example where we make streaming predictions of infectious disease incidence using data from a large database. Copyright © 2017 John Wiley & Sons, Ltd.
| S-EPMC5671383 | biostudies-literature
Project description:Genome-wide association studies (GWASs) have identified several single-nucleotide polymorphisms (SNPs) influencing the risk of Hodgkin's lymphoma (HL) and demonstrated the association of common genetic variation for this type of cancer. Such evidence for inherited genetic risk is also provided by the family history and the very high concordance between monozygotic twins. However, little is known about the genetic and environmental contributions. A common measure for describing the phenotypic variation due to genetics is the heritability. Using GWAS data on 906 HL cases by considering all typed SNPs simultaneously, we have calculated that the common variance explained by SNPs accounts for >35% of the total variation on the liability scale in HL (95% confidence interval 6-62%). These findings are consistent with similar heritability estimates of ∼ 0.40 (95% confidence interval 0.17-0.58) based on Swedish population data. Our estimates support the underlying polygenic basis for susceptibility to HL, and show that heritability based on the population data is somehow larger than heritability based on the genomic data because of the possibility of some missing heritability in the GWAS data. Besides that there is still major evidence for multiple loci causing HL on chromosomes other than chromosome 6 that need to be detected. Because of limited findings in prior GWASs, it seems worth checking for more loci causing susceptibility to HL.
| S-EPMC4795060 | biostudies-other
Project description:Confidence intervals (CIs) depict the statistical uncertainty surrounding evolutionary divergence time estimates. They capture variance contributed by the finite number of sequences and sites used in the alignment, deviations of evolutionary rates from a strict molecular clock in a phylogeny, and uncertainty associated with clock calibrations. Reliable tests of biological hypotheses demand reliable CIs. However, current non-Bayesian methods may produce unreliable CIs because they do not incorporate rate variation among lineages and interactions among clock calibrations properly. Here, we present a new analytical method to calculate CIs of divergence times estimated using the RelTime method, along with an approach to utilize multiple calibration uncertainty densities in dating analyses. Empirical data analyses showed that the new methods produce CIs that overlap with Bayesian highest posterior density intervals. In the analysis of computer-simulated data, we found that RelTime CIs show excellent average coverage probabilities, that is, the actual time is contained within the CIs with a 94% probability. These developments will encourage broader use of computationally efficient RelTime approaches in molecular dating analyses and biological hypothesis testing.
| S-EPMC6984362 | biostudies-literature
Project description:ObjectiveThis study attempted to clarify the applicability of standard error (SE) terms in clinical research when examining the impact of short-term practice effects on cognitive performance via reliable change methodology.MethodThis study compared McSweeney's SE of the estimate (SEest) to Crawford and Howell's SE for prediction of the regression (SEpred) using a developmental sample of 167 participants with either normal cognition or mild cognitive impairment (MCI) assessed twice over 1 week. One-week practice effects in older adults: Tools for assessing cognitive change. Using these SEs, previously published standardized regression-based (SRB) reliable change prediction equations were then applied to an independent sample of 143 participants with MCI.ResultsThis clinical developmental sample yielded nearly identical SE values (e.g., 3.697 vs. 3.719 for HVLT-R Total Recall SEest and SEpred, respectively), and the resultant SRB-based discrepancy z scores were comparable and strongly correlated (r = 1.0, p < .001). Consequently, observed follow-up scores for our sample with MCI were consistently below expectation compared to predictions based on Duff's SRB algorithms.ConclusionsThese results appear to replicate and extend previous work showing that the calculation of the SEest and SEpred from a clinical sample of cognitively intact and MCI participants yields similar values and can be incorporated into SRB reliable change statistics with comparable results. As a result, neuropsychologists utilizing reliable change methods in research investigation (or clinical practice) should carefully balance mathematical accuracy and ease of use, among other factors, when determining which SE metric to use.
| S-EPMC8060987 | biostudies-literature