Bootstrap Methods for Statistical Inference. Part I: Comparative Forecast Verification for Continuous Variables

Gilleland, Eric

doi:10.1175/jtech-d-20-0069.1

Cited by 12 publications

(6 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Estimated 95% CIs using the DM method are compared with estimates using three other methods in Table C1. The commonly used Student's t-distribution approximation (e.g., Gilleland 2020Gilleland , p. 2123 is not recommended in this application since the assumption of serial independence is violated here, and it offers no advantage over the DM method in relation to zero inflated data.…”

Section: Appendix C Confidence Intervals and Hypothesis Testingmentioning

confidence: 99%

“…Circular block bootstrap sampling provides a way to respect serial correlation in the data. Here we used block sizes of length 27 (which is approximately √ 731, and "much longer than the length of dependence, but much shorter than the entire series"; Gilleland 2020), generated 27,000 bootstrap samples and applied the percentile bootstrap method (Gilleland, 2020(Gilleland, , p. 2125) to estimate CIs. The limitation is that bootstrapped samples are drawn from data with a small proportion of non-zero data points.…”

Section: Appendix C Confidence Intervals and Hypothesis Testingmentioning

confidence: 99%

See 1 more Smart Citation

A scoring framework for tiered warnings and multicategorical forecasts based on fixed risk measures

Taggart

Loveday

Griffiths

2022

Quart J Royal Meteoro Soc

View full text Add to dashboard Cite

The use of tiered warnings and multicategorical forecasts are ubiquitous in meteorological operations. Here, a flexible family of scoring functions is presented for evaluating the performance of ordered multicategorical forecasts. Each score has a risk parameter 𝛼, selected for the specific use case, so that it is consistent with a forecast directive based on the fixed threshold probability 1 − 𝛼 (equivalently, a fixed 𝛼-quantile mapping). Each score also has use-case specific weights so that forecasters who accurately discriminate between categorical thresholds are rewarded in proportion to the weight for that threshold. A variation is presented where the penalty assigned to near misses or close false alarms is discounted, which again is consistent with directives based on fixed risk measures. The scores presented provide an alternative to many performance measures currently in use, whose optimal threshold probabilities for forecasting an event typically vary with each forecast case, and in the case of equitable scores are based around sample base rates rather than risk measures suitable for users.

show abstract

Section: Appendix C Confidence Intervals and Hypothesis Testingmentioning

confidence: 99%

Section: Appendix C Confidence Intervals and Hypothesis Testingmentioning

confidence: 99%

A scoring framework for tiered warnings and multicategorical forecasts based on fixed risk measures

Taggart

Loveday

Griffiths

2022

Quart J Royal Meteoro Soc

View full text Add to dashboard Cite

show abstract

“…Predicted values from the RF models were mapped at 25 m pixel resolution using ESRI ArcGIS Pro 2.7.0. Due to heavy-tail distribution of the data, an m-out-of-n bootstrap was used to discern statistical difference between the maps using the R package distillery [62] with Pearson correlation tests for a random set of 1000 subsampled points, bootstrapped over 1000 iterations with resampling.…”

Section: Validation and Mappingmentioning

confidence: 99%

Identifying and Mapping Groups of Protected Area Visitors by Environmental Awareness

Gosal

McMahon

Bowgen

et al. 2021

Land

View full text Add to dashboard Cite

Protected areas worldwide receive billions of visitors annually. The positive impact of nature on health and wellbeing, in addition to providing opportunities for cultural activities such as recreation and aesthetic appreciation, is well documented. Management to reduce negative impacts to biodiversity and conservation aims whilst providing amenities and access to visitors is important. Understanding environmental awareness of visitors and their on-site spatial patterns can assist in making effective management decisions within often constrained resources. However, there is a lack of strategies for site-specific identification and predictive mapping of visitors by environmental awareness. Here, we demonstrate a method to map on-site visitation by latent groups of visitors based on their environmental awareness of on-site issues. On-site surveys and participatory mapping were used to collect data on environmental awareness on bird nesting and spatial visitation patterns in an upland moor in northern England. Latent class analysis (LCA), a structural equation model, was used to discover underlying groups of environmental awareness, with random forest (RF) modelling, a machine learning technique, using a range of on-site predictors (bioclimatic, land cover, elevation, viewshed, and proximity to paths and freshwater) to predict and map visitation across the site by each group. Visitors were segmented into ‘aware’ and ‘ambiguous’ groups and their potential spatial visitation patterns mapped. Our results demonstrate the ability to uncover groups of users by environmental awareness and map their potential visitation across a site using a variety of on-site predictors. Spatial understanding of the movement patterns of differently environmentally aware groups of visitors can assist in efficient targeting of conservation education endeavours (i.e., signage, positioning of staff, monitoring programmes, etc.), therefore maximising their efficacy. Furthermore, we anticipate this method will be of importance to environmental managers and educators when deploying limited resources.

show abstract

“…It should be noted that the probability of occurrence of such an event over a number of years, M, is possibly higher than one might think (cf. Gilleland et al 2017). For example, suppose a home is for sale by a river.…”

Section: Extreme Value Analysismentioning

confidence: 99%

“…The main objective of this paper is to describe new R software (R Core Team 2017), available in the ''extRemes'' (Gilleland and Katz 2016) package, and making use of code from the ''distillery'' (Gilleland 2017) package, for obtaining accurate CI's for extreme values.…”

Section: Introductionmentioning

confidence: 99%

Bootstrap Methods for Statistical Inference. Part II: Extreme-Value Analysis

Gilleland

2020

Journal of Atmospheric and Oceanic Technology

Self Cite

View full text Add to dashboard Cite

This paper is the sequel to a companion paper on bootstrap resampling that reviews bootstrap methodology for making statistical inferences for atmospheric science applications where the necessary assumptions are often not met for the most commonly used resampling procedures. In particular, this sequel addresses extreme-value analysis applications with discussion on the challenges for finding accurate bootstrap methods in this context. New bootstrap code from the R packages distillery and extRemes is introduced. It is further found that one approach for accurate confidence intervals in this setting is not well suited to the case when the random sample’s distribution is not stationary.

show abstract

Bootstrap Methods for Statistical Inference. Part I: Comparative Forecast Verification for Continuous Variables

Cited by 12 publications

References 36 publications

A scoring framework for tiered warnings and multicategorical forecasts based on fixed risk measures

A scoring framework for tiered warnings and multicategorical forecasts based on fixed risk measures

Identifying and Mapping Groups of Protected Area Visitors by Environmental Awareness

Bootstrap Methods for Statistical Inference. Part II: Extreme-Value Analysis

Contact Info

Product

Resources

About