Models are used to explore the global
#foodsystem, incl. implications for
#nutrition, with unclear confidence. In a
#SensitivityAnalysis of a model, variations in the calculated supply of some nutrients led to qualitative changes from sufficient to insufficient global supply:
doi.org/10.1007/s125...Client Challenge
Sensitivity analysis of global food and nutrition modelling - Food Security
Computational models are often used to explore the future of the global food system, including the implications for human nutrition, an essential aspect of sustainability. However, the confidence that can be placed in the outputs of these models is often poorly quantified. Here, a sensitivity analysis of the DELTA Model® - a linear mass balance model calculating global nutrient supply using global and regional food balance sheet, processing, waste, inedible portion, composition, and bioavailability datasets - is conducted. First, a one-at-a-time analysis, varying 4019 underpinning datapoints from the above datasets individually by ± 50% was conducted to identify those with the greatest impact on calculated global nutrient supply. The most influential values from this initial analysis were then carried forward into a multiple value sensitivity analysis, where all possible combinations of ± 50% variations were simulated. Values related to cereals supply, waste, and nutritional value were the most influential on model output, with selenium, cystine, and carbohydrate supply the most sensitive nutrients. When compared to global nutrient requirements, variations in the calculated supply of some nutrients led to qualitative changes from a sufficient global supply to an insufficient supply. These results, while indicative rather than precise estimates of uncertainty, emphasise the critical importance of accurate cereals data in food system models, provide insight on the degree of sensitivity of similar linear models, and should encourage broader application of sensitivity analysis in the field.
Bayesian workflow: Prior determination, predictive checks and sensitivity analyses | Pablo Bernabeu
This post presents a run-through of a Bayesian workflow in R. The content is closely based on Bernabeu (2022), which was in turn based on lots of other references, also cited here.
#statstab #409 Sensitivity Analyses for Unmeasured Confounders
Thoughts: Some assumptions of obs. research are untestable. One way around this is testing what could break your inference.
#causalinference #confounder #collider #bias #sensitivityanalysis #observational
https://link.springer.com/article/10.1007/s40471-022-00308-6
Sensitivity Analyses for Unmeasured Confounders - Current Epidemiology Reports
Purpose of Review This review expands on sensitivity analyses for unmeasured confounding techniques, demonstrating state-of-the-art methods as well as specifying which should be used under various scenarios, depending on the information about a potential unmeasured confounder available to the researcher. Recent Findings Methods to assess how sensitive an observed estimate is to unmeasured confounding have been developed for decades. Recent advancements have allowed for the incorporation of measured confounders in these assessments, updating the methods used to quantify the impact of an unmeasured confounder, whether specified in terms of the magnitude of the effect from a regression standpoint, for example, as a risk ratio, or with respect to the percent of variation in the outcome or exposure explained by the unmeasured confounder. Additionally, single number summaries, such as the E-value or robustness value, have been proposed to allow for ease of computation when less is known about a specific potential unmeasured confounder. Summary This paper aimed to provide methods and tools to implement sensitivity to unmeasured confounder analyses appropriate for various research settings depending on what is known or assumed about a potential unmeasured confounder. We have provided mathematical justification, recommendations, as well as R code to ease the implementation of these methods.
PhD Position Symbolic AI and Reasoning Under Uncertainty
PhD Position Symbolic AI and Reasoning Under Uncertainty
One month left to apply!
If you are looking for a PhD position and are interested in working on probabilistic inference, sensitivity analysis, and decision-making, this might be the job for you! We are looking for candidates with a strong background in Computer Science, and ideally also in Mathematics.
Please apply by 31 August. We're looking forward to reading your application!
https://careers.tudelft.nl/job/Delft-PhD-Position-Symbolic-AI-and-Reasoning-Under-Uncertainty-2628-CD/824585702/
#AcademicJobs
#AcademicMastodon
#GetFediHired
#AcademicJob
#SymbolicAI
#Statistics
#AI
#ConstraintProgramming
#CombinatorialOptimisation
#SensitivityAnalysis
#FormalMethods
#CombinatorialOptimization
#Delft
#TUDelft
#AcademicChatter
PhD Position Symbolic AI and Reasoning Under Uncertainty
PhD Position Symbolic AI and Reasoning Under Uncertainty
PhD Position Symbolic AI and Reasoning Under Uncertainty
PhD Position Symbolic AI and Reasoning Under Uncertainty
💡 Check out this study on key parameters affecting #Antarctic #seaice 🧊 simulation in #Earth system models. With 449 model runs and advanced #sensitivityanalysis, find out how optimizing model parameters can improve sea ice extent and volume simulation!
Read more below👇
https://bit.ly/3ZxZrko
#H2020
Sensitivity of NEMO4.0-SI3 model parameters on sea ice budgets in the Southern Ocean
Abstract. The seasonally dependent Antarctic sea ice concentration
(SIC) budget is well observed and synthesizes many important air–sea–ice
interaction processes. However, it is rarely well simulated in Earth system
models, and means to tune the former are not well understood. In this study,
we investigate the sensitivity of 18 key NEMO4.0-SI3 (Nucleus for
European Modelling of the Ocean coupled with the Sea Ice Modelling
Integrated Initiative) model parameters on modelled SIC and sea ice volume
(SIV) budgets in the Southern Ocean based on a total of 449 model runs and
two global sensitivity analysis methods. We found that the simulated SIC and SIV
budgets are sensitive to ice strength, the thermal conductivity of snow, the
number of ice categories, two parameters related to lateral melting,
ice–ocean drag coefficient and air–ice drag coefficient. An optimized
ice–ocean drag coefficient and air–ice drag coefficient can reduce the
root-mean-square error between simulated and observed SIC budgets by about
10 %. This implies that a more accurate calculation of ice velocity is the
key to optimizing the SIC budget simulation, which is unlikely to be
achieved perfectly by simply tuning the model parameters in the presence of
biased atmospheric forcing. Nevertheless, 10 combinations of
NEMO4.0-SI3 model parameters were recommended, as they could yield
better sea ice extent and SIC budgets than when using the standard values.
New blogpost 🚨 : What do you do if you found a significant result, but your study was underpowered? How reliable is your finding? I discuss Type M and Type S error
https://mzstats.blogspot.com/2023/02/what-not-to-do-with-non-null-results.html
#statistics #frequentist #NHST #pvalue #sensitivityanalysis #falsepositiverisk #rstats
What NOT to do with NON-“null” results – Part III: Underpowered study, but significant result
underpowered, null, statistics
Alexander J. Stein
💧🌏 Greg Cocks
Dr Mircea Zloteanu 🌺🌞🍃
Dr. Anna Latour
PolarRES
Trev