Prediction intervals(PIs)for industrial time series can provide useful guidance for workers.Given that the failure of industrial sensors may cause the missing point in inputs,the existing kernel dynamic Bayesian netwo...Prediction intervals(PIs)for industrial time series can provide useful guidance for workers.Given that the failure of industrial sensors may cause the missing point in inputs,the existing kernel dynamic Bayesian networks(KDBN),serving as an effective method for PIs construction,suffer from high computational load using the stochastic algorithm for inference.This study proposes a variational inference method for the KDBN for the purpose of fast inference,which avoids the timeconsuming stochastic sampling.The proposed algorithm contains two stages.The first stage involves the inference of the missing inputs by using a local linearization based variational inference,and based on the computed posterior distributions over the missing inputs the second stage sees a Gaussian approximation for probability over the nodes in future time slices.To verify the effectiveness of the proposed method,a synthetic dataset and a practical dataset of generation flow of blast furnace gas(BFG)are employed with different ratios of missing inputs.The experimental results indicate that the proposed method can provide reliable PIs for the generation flow of BFG and it exhibits shorter computing time than the stochastic based one.展开更多
In this article, analytical results are obtained apparently for the first time in the literature, for the lower and upper bounds of the roots of quadratic equations when two or all three coefficients a, b, c constitut...In this article, analytical results are obtained apparently for the first time in the literature, for the lower and upper bounds of the roots of quadratic equations when two or all three coefficients a, b, c constitute an interval, with a method called the sign-variation analysis. The results are compared with the parametrization technique offered by Elishakoff and Miglis, and with the solution yielded by minimization and maximization commands of the Maple software. Solutions for some interval word problems are also provided to edulcorate the methodology. This article only focuses on the real roots of those quadratic equations, complex solutions being beyond this investigation.展开更多
In data envelopment analysis (DEA), input and output values are subject to change for several reasons. Such variations differ in their input/output items and their decision-making units (DMUs). Hence, DEA efficiency s...In data envelopment analysis (DEA), input and output values are subject to change for several reasons. Such variations differ in their input/output items and their decision-making units (DMUs). Hence, DEA efficiency scores need to be examined by considering these factors. In this paper, we propose new resampling models based on these variations for gauging the confidence intervals of DEA scores. The first model utilizes past-present data for estimating data variations imposing chronological order weights which are supplied by Lucas series (a variant of Fibonacci series). The second model deals with future prospects. This model aims at forecasting the future efficiency score and its confidence interval for each DMU. We applied our models to a dataset composed of Japanese municipal hospitals.展开更多
Accurately forecasting gasoline volatility is significant for risk management,economic analysis,and option pricing formulas for future contracts.This study proposes a novel interval-valued hierarchical decomposition a...Accurately forecasting gasoline volatility is significant for risk management,economic analysis,and option pricing formulas for future contracts.This study proposes a novel interval-valued hierarchical decomposition and ensemble(IHDE)approach to investigate gasoline price volatility.Our interval-based IHDE method can decompose the complex price process into different components to capture the distinct features of each component,which is helpful for forecasting and analyzing complex price processes.By using interval-valued data,the dynamics of gasoline prices in terms of levels and variations can be fully utilized in this study.Fully utilizing the informational gain of interval-valued data improves forecasting performance.In forecasting weekly gasoline volatility,we document that the proposed IHDE approach outperforms the GARCH,EGARCH,CARR,and ACI models,indicating the importance of capturing features of different frequency components and utilizing the informational gain of interval-valued data for gasoline volatility forecasts.展开更多
Paleoearthquake research represents an essential method for determining recurrence intervals oflarge earthquakes.Reasonable determination of the average recurrence interval and coefficient of variationprovides a cruci...Paleoearthquake research represents an essential method for determining recurrence intervals oflarge earthquakes.Reasonable determination of the average recurrence interval and coefficient of variationprovides a crucial basis for the analysis of the recurrence characteristics of strong earthquakes on intraplatefaults in Chinese mainland.Paleoearthquake data from 145 fault segments of 93 well-studied faults in MainlandChina were collected,organized,and analyzed to discuss the rational estimation of the average recurrenceinterval and coefficient of variation of a strong earthquake occurrence probability model.First,differencesin structural environments were used as a basis to investigate the spatial distribution characteristics of theaverage recurrence intervals of strong earthquakes.The results indicate significant variations in the recurrenceperiods of strong earthquakes in the Sichuan–Yunnan,Xinjiang,North China,and northeastern Qinghai-Tibet Plateau structure zones.The Sichuan–Yunnan structure zone exhibited the shortest average recurrence intervalfor strong earthquakes,which was mainly distributed between 100 and 2000 years,and a relatively high sliprate.The Xinjiang structure zone attained a relatively balanced recurrence interval frequency distribution of1000–4500 years and a moderate slip rate.The North China structure zone showed the lowest slip rate,withthe strong earthquake recurrence interval mainly concentrated between 1000 and 4000 years.The northeastern Qinghai-Tibet Plateau structure zone presented two main frequency peaks in the strong earthquake recurrenceintervals between 1000–3000 years and 3000–5000 years and a relatively high slip rate.The slip rate is a keyfactor influencing the recurrence interval of strong earthquakes,and active faults with high slip rates showshort recurrence intervals.Furthermore,the relationship between fault slip rate,fault type,and the averagerecurrence interval of strong earthquakes was examined.The results indicate a good logarithmic linearrelationship between the fault slip rate and the average recurrence interval of large earthquakes—the higherthe slip rate,the shorter the recurrence interval of strong earthquakes.Fault type also showed a relation to theaverage recurrence interval,with the intervals for various types of active faults gradually increasing in theorder of strike-slip,normal,reverse strike-slip,reverse,and normal strike-slip faults.Second,we calculated theproportions of active faults and various fault types in each structure zone that had a coefficient of variation inrecurrence intervals less than 0.4.The findings reveal that the occurrence of strong earthquakes on most activefaults in Chinese mainland satisfies a quasiperiodic model.The general coefficient of variation across differentstructure zones and fault types ranges between 0.36 and 0.44,which indicates the nonsignificant difference inthe degree of variability in the periodicity of strong earthquake occurrence across various structural zones andfault types.展开更多
基金supported by the National Key Research andDevelopment Program of China(2017YFA0700300)the National Natural Sciences Foundation of China(61533005,61703071,61603069)。
文摘Prediction intervals(PIs)for industrial time series can provide useful guidance for workers.Given that the failure of industrial sensors may cause the missing point in inputs,the existing kernel dynamic Bayesian networks(KDBN),serving as an effective method for PIs construction,suffer from high computational load using the stochastic algorithm for inference.This study proposes a variational inference method for the KDBN for the purpose of fast inference,which avoids the timeconsuming stochastic sampling.The proposed algorithm contains two stages.The first stage involves the inference of the missing inputs by using a local linearization based variational inference,and based on the computed posterior distributions over the missing inputs the second stage sees a Gaussian approximation for probability over the nodes in future time slices.To verify the effectiveness of the proposed method,a synthetic dataset and a practical dataset of generation flow of blast furnace gas(BFG)are employed with different ratios of missing inputs.The experimental results indicate that the proposed method can provide reliable PIs for the generation flow of BFG and it exhibits shorter computing time than the stochastic based one.
文摘In this article, analytical results are obtained apparently for the first time in the literature, for the lower and upper bounds of the roots of quadratic equations when two or all three coefficients a, b, c constitute an interval, with a method called the sign-variation analysis. The results are compared with the parametrization technique offered by Elishakoff and Miglis, and with the solution yielded by minimization and maximization commands of the Maple software. Solutions for some interval word problems are also provided to edulcorate the methodology. This article only focuses on the real roots of those quadratic equations, complex solutions being beyond this investigation.
文摘In data envelopment analysis (DEA), input and output values are subject to change for several reasons. Such variations differ in their input/output items and their decision-making units (DMUs). Hence, DEA efficiency scores need to be examined by considering these factors. In this paper, we propose new resampling models based on these variations for gauging the confidence intervals of DEA scores. The first model utilizes past-present data for estimating data variations imposing chronological order weights which are supplied by Lucas series (a variant of Fibonacci series). The second model deals with future prospects. This model aims at forecasting the future efficiency score and its confidence interval for each DMU. We applied our models to a dataset composed of Japanese municipal hospitals.
基金Supported by National Natural Science Foundation of China(72322016,72073126,71988101)Beijing Natural Science Foundation(9254024)。
文摘Accurately forecasting gasoline volatility is significant for risk management,economic analysis,and option pricing formulas for future contracts.This study proposes a novel interval-valued hierarchical decomposition and ensemble(IHDE)approach to investigate gasoline price volatility.Our interval-based IHDE method can decompose the complex price process into different components to capture the distinct features of each component,which is helpful for forecasting and analyzing complex price processes.By using interval-valued data,the dynamics of gasoline prices in terms of levels and variations can be fully utilized in this study.Fully utilizing the informational gain of interval-valued data improves forecasting performance.In forecasting weekly gasoline volatility,we document that the proposed IHDE approach outperforms the GARCH,EGARCH,CARR,and ACI models,indicating the importance of capturing features of different frequency components and utilizing the informational gain of interval-valued data for gasoline volatility forecasts.
基金Funded by the National Key Research and Development Program(2022YFC3003502).
文摘Paleoearthquake research represents an essential method for determining recurrence intervals oflarge earthquakes.Reasonable determination of the average recurrence interval and coefficient of variationprovides a crucial basis for the analysis of the recurrence characteristics of strong earthquakes on intraplatefaults in Chinese mainland.Paleoearthquake data from 145 fault segments of 93 well-studied faults in MainlandChina were collected,organized,and analyzed to discuss the rational estimation of the average recurrenceinterval and coefficient of variation of a strong earthquake occurrence probability model.First,differencesin structural environments were used as a basis to investigate the spatial distribution characteristics of theaverage recurrence intervals of strong earthquakes.The results indicate significant variations in the recurrenceperiods of strong earthquakes in the Sichuan–Yunnan,Xinjiang,North China,and northeastern Qinghai-Tibet Plateau structure zones.The Sichuan–Yunnan structure zone exhibited the shortest average recurrence intervalfor strong earthquakes,which was mainly distributed between 100 and 2000 years,and a relatively high sliprate.The Xinjiang structure zone attained a relatively balanced recurrence interval frequency distribution of1000–4500 years and a moderate slip rate.The North China structure zone showed the lowest slip rate,withthe strong earthquake recurrence interval mainly concentrated between 1000 and 4000 years.The northeastern Qinghai-Tibet Plateau structure zone presented two main frequency peaks in the strong earthquake recurrenceintervals between 1000–3000 years and 3000–5000 years and a relatively high slip rate.The slip rate is a keyfactor influencing the recurrence interval of strong earthquakes,and active faults with high slip rates showshort recurrence intervals.Furthermore,the relationship between fault slip rate,fault type,and the averagerecurrence interval of strong earthquakes was examined.The results indicate a good logarithmic linearrelationship between the fault slip rate and the average recurrence interval of large earthquakes—the higherthe slip rate,the shorter the recurrence interval of strong earthquakes.Fault type also showed a relation to theaverage recurrence interval,with the intervals for various types of active faults gradually increasing in theorder of strike-slip,normal,reverse strike-slip,reverse,and normal strike-slip faults.Second,we calculated theproportions of active faults and various fault types in each structure zone that had a coefficient of variation inrecurrence intervals less than 0.4.The findings reveal that the occurrence of strong earthquakes on most activefaults in Chinese mainland satisfies a quasiperiodic model.The general coefficient of variation across differentstructure zones and fault types ranges between 0.36 and 0.44,which indicates the nonsignificant difference inthe degree of variability in the periodicity of strong earthquake occurrence across various structural zones andfault types.
