EX2 Structural Uncertainty in Survival Extrapolation: Exploring the IMPACT of FOUR MODEL Averaging Methods and Adjusting for DATA Maturity

Abstract To study the impact of training sample heterogeneity on the performance of Bayesian model averaging (BMA), two BMA experiments were performed on probabilistic quantitative precipitation forecasts (PQPFs) in the northern China region in July and August of 2010 generated from an 11-member short-range ensemble forecasting system. One experiment, as in many conventional BMA studies, used an overall training sample that consisted of all available cases in the training period, while the second experiment used stratified sampling BMA by first dividing all available training cases into subsamples according to their ensemble spread, and then performing BMA on each subsample. The results showed that ensemble spread is a good criterion to divide ensemble precipitation cases into subsamples, and that the subsamples have different statistical properties. Pooling the subsamples together forms a heterogeneous overall sample. Conventional BMA is incapable of interpreting heterogeneous samples, and produces unreliable PQPF. It underestimates the forecast probability at high-threshold PQPF and local rainfall maxima in BMA percentile forecasts. BMA with stratified sampling according to ensemble spread overcomes the problem reasonably well, producing sharper predictive probability density functions and BMA percentile forecasts, and more reliable PQPF than the conventional BMA approach. The continuous ranked probability scores, Brier skill scores, and reliability diagrams of the two BMA experiments were examined for all available forecast days, along with a logistic regression experiment. Stratified sampling BMA outperformed the raw ensemble and conventional BMA in all verifications, and also showed better skill than logistic regression in low-threshold forecasts.

Download Full-text

Impact of GCM structure uncertainty on hydrological processes in an arid area of China

Hydrology Research ◽

10.2166/nh.2017.227 ◽

2017 ◽

Vol 49 (3) ◽

pp. 893-907 ◽

Cited By ~ 11

Author(s):

Gonghuan Fang ◽

Jing Yang ◽

Yaning Chen ◽

Zhi Li ◽

Philippe De Maeyer

Keyword(s):

Climate Change ◽

General Circulation ◽

Assessment Tool ◽

Circulation Model ◽

Hydrological Processes ◽

Future Climate Change ◽

Management Decision ◽

Structural Uncertainty ◽

Rcp Scenarios ◽

The Impact

Abstract Quantifying the uncertainty sources in assessment of climate change impacts on hydrological processes is helpful for local water management decision-making. This paper investigated the impact of the general circulation model (GCM) structural uncertainty on hydrological processes in the Kaidu River Basin. Outputs of 21 GCMs from the Coupled Model Intercomparison Project Phase 5 (CMIP5) under two representative concentration pathway (RCP) scenarios (i.e., RCP4.5 and RCP8.5), representing future climate change under uncertainty, were first bias-corrected using four precipitation and three temperature methods and then used to force a well-calibrated hydrological model (the Soil and Water Assessment Tool, SWAT) in the study area. Results show that the precipitation will increase by 3.1%–18% and 7.0%–22.5%, the temperature will increase by 2.0 °C–3.3 °C and 4.2 °C–5.5 °C and the streamflow will change by −26% to 3.4% and −38% to −7% under RCP4.5 and RCP8.5, respectively. Timing of snowmelt will shift forward by approximately 1–2 months for both scenarios. Compared to RCPs and bias correction methods, GCM structural uncertainty contributes most to streamflow uncertainty based on the standard deviation method (55.3%) while it is dominant based on the analysis of variance approach (94.1%).

Download Full-text

Determinants of Inflation in Iran Based on Bayesian Model Averaging (BMA)

International Letters of Social and Humanistic Sciences ◽

10.18052/www.scipress.com/ilshs.43.10 ◽

2014 ◽

Vol 43 ◽

pp. 10-17

Author(s):

Mohsen Mehrara ◽

Arezoo Ghazanfari ◽

Motahareh Alsadat Majdzadeh

Keyword(s):

Exchange Rate ◽

Bayesian Model ◽

Bayesian Model Averaging ◽

Model Averaging ◽

Energy Price ◽

Market Exchange ◽

Explanatory Variables ◽

Long Run ◽

The Impact ◽

Money Velocity

Due to the important influence of inflation on macro-economic variables, researchers pay tremendous amount of attention to its determinants. Accordingly, in the following research, the impact of 13 variables on inflation during the period of 1338-1391 by using Bayesian Model Averaging (BMA) method has been investigated for Iran economy. The ranking of the 13 explanatory variables are obtained based on the probability of their inclusion in model. The results show that the energy price and money imbalance (lagged ratio of money to nominal output) have expected and positive effect on inflation rate with a probability of 100 % and they are considered as the key explanatory variables in inflation equation. The energy price, money imbalance, money growth and market exchange rate growth have the first to fourth rank respectively. The influence of the production growth is not significant on the inflation in the short-run but it gradually influences the inflation through money imbalance channel in the long-run. In addition, most of the disinflation effects due to decrease in money supply will appear with delay. These results imply the dominance of monetary variables on inflation with cost push factors not having important impacts on prices. Also, oil revenue and imports influence the inflation through exchange rate channel, production and money velocity.

Download Full-text

What Influences Demand for Legal Aid? Logit Analysis with Use of Classical Model Averaging

Przegląd Statystyczny ◽

10.5604/01.3001.0014.0526 ◽

2019 ◽

Vol 65 (1) ◽

pp. 55-82

Author(s):

Waldemar Florczak ◽

Wojciech Grabowski

Keyword(s):

Logit Model ◽

Quantitative Research ◽

Classical Model ◽

Relevant Literature ◽

Occupational Status ◽

Social Activity ◽

Model Averaging ◽

Logit Analysis ◽

Legal Aid ◽

The Impact

The aim of this paper to identify, by means of relevant literature survey, and then to quantify – using representative panel of individual questionnaire data and a logit model – the impact of relevant factors affecting demand for legal aid in Poland. The set of explanatory variables contains objective factors – such as income, age, education, gender, marital status, place of residence, occupational status – as well as subjective ones, such as personal attitude towards law, knowledge of law, social capital or social activity. It follows from the results obtained on a representative sample of adult Poles that the number of factors influencing the occurrence of a legal problem is large, the factors themselves being beyond the scope of direct and intentional impact of the state. Thereby, it seems reasonable to allocate the funds devoted to the funding and functioning of the reformed legal aid system in Poland on the basis of the population size criterion. In view of the relative scarcity of quantitative research into the issues raised in the paper frequentist model averaging method has been also used to confirm/reject the conclusions draw on the basis of the logit model. However, this has not altered the afore-mentioned conclusions in any significant way.

Download Full-text

Barrier Option Pricing with Model Averaging Methods under Local Volatility Models

Industrial Engineering & Management Systems ◽

10.7232/iems.2011.10.1.084 ◽

2011 ◽

Vol 10 (1) ◽

pp. 84-94 ◽

Cited By ~ 2

Author(s):

Nam-Hyoung Kim ◽

Kyu-Hwan Jung ◽

Jae-Wook Lee ◽

Gyu-Sik Han

Keyword(s):

Option Pricing ◽

Model Averaging ◽

Barrier Option ◽

Local Volatility ◽

Averaging Methods ◽

Volatility Models ◽

Barrier Option Pricing

Download Full-text

Radiative forcing and climate metrics for ozone precursor emissions: the impact of multi-model averaging

Atmospheric Chemistry and Physics Discussions ◽

10.5194/acpd-14-27195-2014 ◽

2014 ◽

Vol 14 (19) ◽

pp. 27195-27231

Author(s):

C. R. MacIntosh ◽

K. P. Shine ◽

W. J. Collins

Keyword(s):

Radiative Forcing ◽

Transport Model ◽

Task Force ◽

Mean Field ◽

Model Averaging ◽

Chemical Transport Model ◽

Ensemble Mean ◽

The Mean ◽

The Individual ◽

The Impact

Abstract. Multi-model ensembles are frequently used to assess understanding of the response of ozone and methane lifetime to changes in emissions of ozone precursors such as NOx, VOC and CO. When these ozone changes are used to calculate radiative forcing (RF) (and climate metrics such as the global warming potential (GWP) and global temperature potential (GTP)) there is a methodological choice, determined partly by the available computing resources, as to whether the mean ozone (and methane lifetime) changes are input to the radiation code, or whether each model's ozone and methane changes are used as input, with the average RF computed from the individual model RFs. We use data from the Task Force on Hemispheric Transport of Air Pollution Source-Receptor global chemical transport model ensemble to assess the impact of this choice for emission changes in 4 regions (East Asia, Europe, North America and South Asia). We conclude that using the multi-model mean ozone and methane responses is accurate for calculating the mean RF, with differences up to 0.6% for CO, 0.7% for VOC and 2% for NOx. Differences of up to 60% for NOx 7% for VOC and 3% for CO are introduced into the 20 year GWP as a result of the exponential decay terms, with similar values for the 20 years GTP. However, estimates of the SD calculated from the ensemble-mean input fields (where the SD at each point on the model grid is added to or subtracted from the mean field) are almost always substantially larger in RF, GWP and GTP metrics than the true SD, and can be larger than the model range for short-lived ozone RF, and for the 20 and 100 year GWP and 100 year GTP. We find that the effect is generally most marked for the case of NOx emissions, where the net effect is a smaller residual of terms of opposing signs. For example, the SD for the 20 year GWP is two to three times larger using the ensemble-mean fields than using the individual models to calculate the RF. Hence, while the average of multi-model fields are appropriate for calculating mean RF, GWP and GTP, they are not a reliable method for calculating the uncertainty in these fields, and in general overestimate the uncertainty.

Download Full-text

Incorporating aleatory and epistemic uncertainties in the modelling of construction duration

Engineering Construction & Architectural Management ◽

10.1108/ecam-06-2019-0304 ◽

2020 ◽

Vol 27 (9) ◽

pp. 2199-2219

Author(s):

Hassan Adaviriku Ahmadu ◽

Ahmed Doko Ibrahim ◽

Yahaya Makarfi Ibrahim ◽

Kulomri Jipato Adogbo

Keyword(s):

Construction Projects ◽

Bayesian Model Averaging ◽

Predictive Accuracy ◽

Tertiary Education ◽

Model Averaging ◽

Building Construction ◽

Percentage Error ◽

Content Type ◽

The Impact ◽

Matlab Programming

PurposeThis study aims to develop a model which incorporates the impact of both aleatory and epistemic uncertainties into construction duration predictions, in a manner that is consistent with the nature/quality of information available about various factors which bring about uncertainties.Design/methodology/approachData relating to 178 completed Tertiary Education Trust Fund (TETfund) building construction projects were obtained from construction firms via questionnaire survey. Using 90% of the data, the model was developed in the form of a hybrid-based algorithm implemented through a suitable user-friendly graphical user interface (GUI) using MATLAB programming language. Bayesian model averaging, Monte Carlo simulation and fuzzy logic were the statistical methods used for the algorithm development, prior to its GUI implementation in MATLAB. Using the remaining 10% data, the model's predictive accuracy was assessed via the independent samples t-test and the mean absolute percentage error (MAPE).FindingsThe developed model's predictions were found not statistically different from those of actual duration estimates in the 10% test data, with a MAPE of just 2%. This suggests that the model's ability to incorporate both aleatory and epistemic uncertainties improves accuracy of duration predictions made using it.Research limitations/implicationsThe model was developed using a particular type of building projects (TETfund building construction projects), and so its use is limited to projects with characteristics similar to those used for its development.Practical implicationsThe developed model's predictions are expected to serve as a useful basis for consultancy firms and contractor organisations to make more realistic schedules and benchmark measures of construction period, thereby facilitating effective planning and successful execution of construction projects.Originality/valueThe study presented a model which permits combined manipulation of aleatory and epistemic uncertainties, hence ensuring a more realistic incorporation of uncertainty into construction duration predictions.

Download Full-text

CALIPSO lidar level 3 aerosol profile product: version 3 algorithm design

Atmospheric Measurement Techniques ◽

10.5194/amt-11-4129-2018 ◽

2018 ◽

Vol 11 (7) ◽

pp. 4129-4152 ◽

Cited By ~ 29

Author(s):

Jason L. Tackett ◽

David M. Winker ◽

Brian J. Getzewich ◽

Mark A. Vaughan ◽

Stuart A. Young ◽

...

Keyword(s):

Algorithm Design ◽

Global Ocean ◽

Orthogonal Polarization ◽

Aerosol Extinction ◽

Averaging Methods ◽

Global Land ◽

Level 3 ◽

Classification Information ◽

The Impact ◽

Level 2

Abstract. The CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations) level 3 aerosol profile product reports globally gridded, quality-screened, monthly mean aerosol extinction profiles retrieved by CALIOP (the Cloud-Aerosol Lidar with Orthogonal Polarization). This paper describes the quality screening and averaging methods used to generate the version 3 product. The fundamental input data are CALIOP level 2 aerosol extinction profiles and layer classification information (aerosol, cloud, and clear-air). Prior to aggregation, the extinction profiles are quality-screened by a series of filters to reduce the impact of layer detection errors, layer classification errors, extinction retrieval errors, and biases due to an intermittent signal anomaly at the surface. The relative influence of these filters are compared in terms of sample rejection frequency, mean extinction, and mean aerosol optical depth (AOD). The “extinction QC flag” filter is the most influential in preventing high-biases in level 3 mean extinction, while the “misclassified cirrus fringe” filter is most aggressive at rejecting cirrus misclassified as aerosol. The impact of quality screening on monthly mean aerosol extinction is investigated globally and regionally. After applying quality filters, the level 3 algorithm calculates monthly mean AOD by vertically integrating the monthly mean quality-screened aerosol extinction profile. Calculating monthly mean AOD by integrating the monthly mean extinction profile prevents a low bias that would result from alternately integrating the set of extinction profiles first and then averaging the resultant AOD values together. Ultimately, the quality filters reduce level 3 mean AOD by −24 and −31 % for global ocean and global land, respectively, indicating the importance of quality screening.

Download Full-text