Deviation, Risk, and Error Measures

2013 ◽  
pp. 19-31
Author(s):  
Michael Zabarankin ◽  
Stan Uryasev
Keyword(s):  
Error Measures

scholarly journals The accuracy of alternative GDP growth forecasts: Do they represent a credible alternative to the official ones?

2015 ◽  
Vol 15 (3) ◽  
pp. 291-304
Author(s):  
Jiří Šindelář
Keyword(s):  
Trade Unions ◽  
Time Frame ◽  
Gdp Growth ◽  
Gradual Improvement ◽  
Error Measures ◽  
National Bank ◽  
Ministry Of Finance ◽  
Governmental Institutions ◽  
High Level ◽  
Growth Forecasts

Abstract The paper deals with the accuracy of the real GDP growth forecasts produced by two Czech non-governmental institutions: the Czech-Moravian Confederation of Trade Unions (CMKOS) and the Czech Banking Association (CBA) in the years 2007-2014 and 2011-2014 respectively. Utilizing a research method composed of simple (AFE), scale-dependent (RMSE) as well as relative (MASE) error measures, we found out that (i) CMKOS predictions achieved a lower forecasting error on average, beginning with a notable overestimation in the first turnover point from growth to decline (2008-2009), yet followed by gradual improvement resulting in superior accuracy over set benchmarks (Ministry of Finance, Czech National Bank, OECD) in later years (2010-2014). The CBA predictions, on the other hand, exhibited (ii) a high level of interconnection with official bodies (MF, CNB), but with overall inferior forecasting accuracy, despite the shorter time frame (2011-2014). Overall, the study suggests that of the two surveyed non-governmental bodies, only CMKOS forecasts represent a viable alternative to the official predictions published by the Ministry of Finance or the Czech National Bank, as CBA forecasts were found to be a less accurate satellite of these bodies.

Limits of a simplified controller design based on integral plus dead time models

2018 ◽  
Vol 232 (6) ◽  
pp. 728-741 ◽  
Author(s):  
Mikuláš Huba ◽  
Igor Bélai
Keyword(s):  
Dead Time ◽  
Controller Design ◽  
Absolute Error ◽  
Higher Order ◽  
Smith Predictor ◽  
Proportional Integral ◽  
Overall Design ◽  
Error Measures ◽  
Integral Controller ◽  
Proportional Integral Controller

This article presents design and evaluation of filtered proportional–integral controllers and filtered Smith predictor–inspired constrained dead time compensators. Both are based on the integral plus dead time and on the first-order time delayed plant models. They are compared as for tuning simplicity, robustness and noise attenuation. Such a comparison, which presents a robustness test regarding the importance of the internal plant feedback approximation, may be carried out by performance measures built on deviations of the input and output transient responses from their ideal shapes. When combined with integral of absolute error measures of both solution types with the disturbance responses set as nearly equivalent, we can see that the filtered Smith predictor setpoint responses may be significantly faster than the filtered proportional–integral controller responses, more robust and, using higher-order filters, also sufficiently smooth. Furthermore, tuning of the possibly higher-order filters for filtered Smith predictor is simpler. Its overall design is more transparent and straightforward with respect to the control constraints, where the filtered Smith predictor requires some additional anti-windup measures.

scholarly journals Quantitative Assessment of the Influence of Tensile Softening of Concrete in Beams under Bending by Numerical Simulations with XFEM and Cohesive Cracks

Materials ◽  
2022 ◽  
Vol 15 (2) ◽  
pp. 626
Author(s):  
Ireneusz Marzec ◽  
Jerzy Bobiński
Keyword(s):  
Experimental Data ◽  
Numerical Simulations ◽  
Extended Finite Element ◽  
Three Point Bending ◽  
Material Parameters ◽  
Error Measures ◽  
Rational Bézier Curve ◽  
Initial Fracture ◽  
Experimental Values ◽  
Tensile Softening

Results of the numerical simulations of the size effect phenomenon for concrete in comparison with experimental data are presented. In-plane geometrically similar notched and unnotched beams under three-point bending are analyzed. EXtended Finite Element Method (XFEM) with a cohesive softening law is used. Comprehensive parametric study with the respect to the tensile strength and the initial fracture energy is performed. Sensitivity of the results with respect to the material parameters and the specimen geometry is investigated. Three different softening laws are examined. First, a bilinear softening definition is utilized. Then, an exponential curve is taken. Finally, a rational Bezier curve is tested. An ambiguity in choosing material parameters and softening curve definitions is discussed. Numerical results are compared with experimental outcomes recently reported in the literature. Two error measures are defined and used to quantitatively assess calculated maximum forces (nominal strengths) in comparison with experimental values as a primary criterion. In addition, the force—displacement curves are also analyzed. It is shown that all softening curves produce results consistent with the experimental data. Moreover, with different softening laws assumed, different initial fracture energies should be taken to obtain proper results.

scholarly journals An Order-Independent Algorithm for Learning Chain Graphs

2021 ◽  
Vol 34 (1) ◽  
Author(s):  
Mohammad Ali Javidian ◽  
Marco Valtorta ◽  
Pooyan Jamshidi
Keyword(s):  
Directed Acyclic Graphs ◽  
High Dimensional ◽  
P Values ◽  
Order Dependence ◽  
Error Measures ◽  
Acyclic Graphs ◽  
Simulation Results ◽  
Improved Performance ◽  
Low Dimensional ◽  
Chain Graphs

LWF chain graphs combine directed acyclic graphs and undirected graphs. We propose a PC-like algorithm, called PC4LWF, that finds the structure of chain graphs under the faithfulness assumption to resolve the problem of scalability of the proposed algorithm by Studeny (1997). We prove that PC4LWF is order dependent, in the sense that the output can depend on the order in which the variables are given. This order dependence can be very pronounced in high-dimensional settings. We propose two modifications of the PC4LWF algorithm that remove part or all of this order dependence. Simulation results with different sample sizes, network sizes, and p-values demonstrate the competitive performance of the PC4LWF algorithms in comparison with the LCD algorithm proposed by Ma et al. (2008) in low-dimensional settings and improved performance (with regard to error measures) in high-dimensional settings.

Sign in / Sign up

Export Citation Format

Share Document