Big data: Some statistical issues

Cox, D. R.; Kartsonaki, Christiana; Keogh, Ruth H.

doi:10.1016/j.spl.2018.02.015

Cited by 50 publications

(29 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Moreover, in contrast to several statistical methodologies for probabilistic prediction, such as the Bayesian methodology by Tyralis and Koutsoyiannis (2014), a well-known drawback of flexible statistical learning models for quantile prediction is their inappropriateness for modelling long-range dependence (see also Cox et al 2018).…”

Section: Summary Discussion and Conclusionmentioning

confidence: 99%

Quantification of predictive uncertainty in hydrological modelling by harnessing the wisdom of the crowd: Methodology development and investigation using toy models

Papacharalampous

Koutsoyiannis

Montanari

2020

Advances in Water Resources

View full text Add to dashboard Cite

We introduce an ensemble learning post-processing methodology for probabilistic hydrological modelling. This methodology generates numerous point predictions by applying a single hydrological model, yet with different parameter values drawn from the respective simulated posterior distribution. We call these predictions "sister predictions". Each sister prediction extending in the period of interest is converted into a probabilistic prediction using information about the hydrological model's errors. This information is obtained from a preceding period for which observations are available, and is exploited using a flexible quantile regression model.All probabilistic predictions are finally combined via simple quantile averaging to produce the output probabilistic prediction. The idea is inspired by the ensemble learning methods originating from the machine learning literature. The proposed methodology offers larger robustness in performance than basic post-processing methodologies using a single hydrological point prediction. It is also empirically proven to "harness the wisdom of the crowd" in terms of average interval score, i.e., the obtained quantile predictions score no worse -usually better− than the average score of the combined individual predictions. This proof is provided within toy examples, which can be used for gaining insight on how the methodology works and under which conditions it can optimally convert point hydrological predictions to probabilistic ones.A large-scale hydrological application is made in a companion paper.

show abstract

Section: Summary Discussion and Conclusionmentioning

confidence: 99%

Quantification of predictive uncertainty in hydrological modelling by harnessing the wisdom of the crowd: Methodology development and investigation using toy models

Papacharalampous

Koutsoyiannis

Montanari

2020

Advances in Water Resources

View full text Add to dashboard Cite

show abstract

“…5 While P values determine the likelihood that a study's results were due to chance or not, these values are often erroneously used as a litmus test to indicate the importance of findings. 6 For example, while the ARRIVE trial results did not reach the preset level of statistical significance for neonatal outcomes, elective labor induction at 39 weeks' gestation was associated with a trend toward a lower stillbirth rate, even though the median gestational age at birth was only 4 days different between the 2 groups. While this outcome was not statistically significant by a small margin, it is consistent with other research in diverse populations, and it may be clinically significant for women and their health care providers, depending on a woman's individual circumstances.…”

mentioning

confidence: 94%

“…When evaluating the results of studies, it is important to distinguish between statistical and clinical significance, as Barger discusses in a commentary in this issue . While P values determine the likelihood that a study's results were due to chance or not, these values are often erroneously used as a litmus test to indicate the importance of findings . For example, while the ARRIVE trial results did not reach the preset level of statistical significance for neonatal outcomes, elective labor induction at 39 weeks’ gestation was associated with a trend toward a lower stillbirth rate, even though the median gestational age at birth was only 4 days different between the 2 groups.…”

mentioning

confidence: 99%

“…Large data sets can be beneficial as they can gather data from whole populations; however, traditional statistical techniques are less valuable in assessing these large data sets. New and advanced statistical techniques estimate relationships among events and outcomes with greater precision . This Journal has featured some of these techniques in previous issues, including Carlson et al's use of propensity‐score matching and Bovbjerg et al's study using Cox proportional‐hazards models .…”

mentioning

confidence: 99%

“…New and advanced statistical techniques estimate relationships among events and outcomes with greater precision. 6 This Journal has featured some of these techniques in previous issues, including Carlson et al's use of propensity-score matching 15 and Bovbjerg et al's study using Cox proportional-hazards models. 16 In this issue, Ellen Tilden, CNM, PhD, and Jon Snowden, PhD, explain some new statistical approaches to study maternity care in a 3-part series.…”

mentioning

confidence: 99%

See 2 more Smart Citations