Proceedings of the 2003 International Conference on Machine Learning and Cybernetics (IEEE Cat. No.03EX693), 2000
Simulation can provide insight to the behavior of a complex queueing system by identifying the re... more Simulation can provide insight to the behavior of a complex queueing system by identifying the response surface of several performance measures such as delays and backlogs. However, simulations of large systems are expensive both in terms of CPU time and use of available resources (e.g. processors). Thus, it is of paramount importance to carefully select the inputs of simulation in order to adequately capture the underlying response surface of interest and at the same time minimize the required number of simulation runs. In this study, we present a methodological framework for designing efficient simulations for complex networks. Our approach works in sequential and combines the methods of CART (Classification And Regression Trees) and the design of experiments. A generalized switch model is used to illustrate the proposed methodology and some useful applications are described.
2006 11th Intenational Workshop on Computer-Aided Modeling, Analysis and Design of Communication Links and Networks, 2000
Switched Processing Systems (SPS) capture the essence of a fundamental resource allocation proble... more Switched Processing Systems (SPS) capture the essence of a fundamental resource allocation problem in many modern communication, computer and manufacturing systems involving heterogeneous processors and multiple classes of job traffic flows. Recently, increased attention has been paid to the issue of improving quality of service (QoS) performance in terms of delays and backlogs of the associated scheduling policies, rather than simply maximizing the system's throughput. In this study, we investigate a class of throughput maximizing scheduling policies called MaxProduct policies. The objective is that through a use of dynamically changing "optimal" queue weights, the corresponding QoS performance measure -e.g. the average system delay-can be significantly improved. The proposed approach involves a statistical smoothing technique for tracking the system's workload and utilizes the result of how the MaxProduct policies drain out an initially placed workload in the shortest possible time. It is further shown that the proposed modification of the MaxProduct policy, achieves maximum throughput without requiring knowledge of the incoming traffic's statistics. The scheduling policy is illustrated on a small SPS subject to different types of input traffic.
We provide a simple proof of Wald's second identity for a class of problems that can be formulate... more We provide a simple proof of Wald's second identity for a class of problems that can be formulated as a fair coin-tossing game. The identity provides a use-ful technique for computing the expected stopping time and allows us to explore the second-order moment of the so-called heads-minus-tails distribution. We also introduce some interesting applications related to this simple identity.
There is increased interest in deploying charging station infrastructure for electric vehicles, d... more There is increased interest in deploying charging station infrastructure for electric vehicles, due to the increasing adoption of such vehicles to reduce emissions. However, there are a number of key challenges for providing high quality of service to such vehicles, stemming from technological reasons. One of them is due to the relative slow charging times and the other is due to the relative limited battery range. Hence, developing efficient routing strategies of electric vehicles requesting charging to stations that have available charging resources is an important component of the infrastructure. In this work, we propose a queueing modeling framework for the problem at hand and develop such routing strategies that optimize a performance metric related to vehicles' sojourn time in the system. By incorporating appropriate weights into the well-known dynamic routing discipline "Join-the-Shortest-Queue", we show that the proposed routing strategies not only do they maximize the queueing system's throughput, but also significantly mitigate the vehicle's sojourn time. The strategies are also adaptive in nature and responsive to changes in the speed of charging at the stations, the distribution of the vehicles' point of origin when requesting service, the traffic congestion level and the vehicle speed; all the above are novel aspects and compatible with the requirements of a modern electric vehicle charging infrastructure.
2007 IEEE International Conference on Communications, 2007
Switched Processing Systems (SPS) represent a canonical model for many areas of applications of c... more Switched Processing Systems (SPS) represent a canonical model for many areas of applications of communication, computer and manufacturing systems. They are characterized by flexible, interdependent service capabilities and multiple classes of job traffic flows. Recently, increased attention has been paid to the issue of improving quality of service (QoS) performance in terms of delays and backlogs of the associated scheduling policies, rather than simply maximizing the system's throughput. In this study, we investigate a measurement based dynamic service allocation policy that significantly improves performance with respect to delay metrics. The proposed policy solves a linear program at selected points in time that are in turn determined by a monitoring strategy that detects 'significant' changes in the intensities of the input processes. The proposed strategy is illustrated on a small SPS subject to different types of input traffic.
Genetic/transcriptional regulatory interactions are shown to predict partial components of signal... more Genetic/transcriptional regulatory interactions are shown to predict partial components of signaling pathways, which have been recognized as vital to complex human diseases. Both activator (A) and repressor (R) are known to coregulate their common target gene (T ). proposed to model this coregulation by a fixed second order response surface (called the RS algorithm), in which T is a function of A, R, and AR. Unfortunately, the RS algorithm did not result in a sufficient number of genetic interactions (GIs) when it was applied to a group of 51 yeast genes in a pilot study.Thus, we propose a data-driven second order model (DDSOM), an approximation to the non-linear transcriptional interactions, to infer genetic and transcriptional regulatory interactions. For each triplet of genes of interest (A, R, andT ), we regress the expression ofT at time t + 1 on the expression of A, R, and AR at time t. Next, these well-fitted regression models (viewed as points in R 3 ) are collected, and the center of these points is used to identify triples of genes having the A-R-T relationship or GIs. The DDSOM and RS algorithms are first compared on inferring transcriptional compensation interactions of a group of yeast genes in DNA synthesis and DNA repair using microarray gene expression data; the DDSOM algorithm results in higher modified true positive rate (about 75%) than that of the RS algorithm, checked against quantitative RT-polymerase chain reaction results.These validated GIs are reported, among which some coincide with certain interactions in DNA repair and genome instability pathways in yeast. This suggests that the DDSOM algorithm has potential to predict pathway components. Further, both algorithms are applied to predict transcriptional regulatory interactions of 63 yeast genes. Checked against the known transcriptional regulatory interactions queried from TRANSFAC, the proposed also performs better than the RS algorithm.
The identification of causal effects between two groups of time series has been an important topi... more The identification of causal effects between two groups of time series has been an important topic in a wide range of applications such as economics, engineering, medicine, neuroscience, and biology. In this paper, a simplified causal relationship (called trimmed Granger causality) based on the context of Granger causality and vector autoregressive (VAR) model is introduced. The idea is to characterize a subset of "important variables" for both groups of time series so that the underlying causal structure can be presented based on minimum variable information. When the VAR model is specified, explicit solutions are provided for the identification of important variables. When the parameters of the VAR model are unknown, an efficient statistical hypothesis testing procedure is introduced to estimate the solution. An example representing the stock indices of different countries is used to illustrate the proposed methods. In addition, a simulation study shows that the proposed methods significantly outperform the Lasso-type methods in terms of the accuracy of characterizing the simplified causal relationship.
Switched Processing Systems (SPS) represent canonical models for many communication and computer ... more Switched Processing Systems (SPS) represent canonical models for many communication and computer systems. Over the years, much research has been devoted to developing the best scheduling policies to optimize the various performance metrics of interest. These policies have mostly originated from the well-known MaxWeight discipline, which at any point in time switches the system into the service mode possessing "maximal matching" with the system state (e.g., queue-length, workload, etc.). However, for simplicity it is often assumed that the switching times between service modes are "negligible"-but this proves to be impractical in some applications. In this study, we propose a new scheduling strategy (called the Dynamic Cone Policy) for SPS, which includes substantial service-mode switching times. The goal is to maximize throughput and maintain system stability under fairly mild stochastic assumptions. For practical purposes, an extended scheduling strategy (called the Practical Dynamic Cone Policy) is developed to reduce the computational complexity of the Dynamic Cone Policy and at the same time mitigate job delay. A simulation study shows that the proposed practical policy clearly outperforms another throughput-maximizing policy called BatchAdapt, both in terms of the average and the 95th percentile of job delay for various types of input traffic.
Simulation can provide insight to the behavior of a complex queueing system by identifying the re... more Simulation can provide insight to the behavior of a complex queueing system by identifying the response surface of several performance measures such as delays and backlogs. However, simulations of large systems are expensive both in terms of CPU time ...
Journal of Statistical Planning and Inference, 2012
We consider the bandit problem with an infinite number of Bernoulli arms, of which the unknown pa... more We consider the bandit problem with an infinite number of Bernoulli arms, of which the unknown parameters are assumed to be i.i.d. random variables with a common distribution F. Our goal is to construct optimal strategies of choosing ''arms'' so that the expected long-run failure rate is minimized. We first review a class of strategies and establish their asymptotic properties when F is known. Based on the results, we propose a new strategy and prove that it is asymptotically optimal when F is unknown. Finally, we show that the proposed strategy performs well for a number of simulation scenarios.
Journal of Statistical Computation and Simulation, 2011
ABSTRACT In this article, we describe various well-known Dirichlet generation algorithms and eval... more ABSTRACT In this article, we describe various well-known Dirichlet generation algorithms and evaluate their performance in terms of the following criteria: (i) computer generation time, (ii) sensitivity, and (iii) goodness of fit. In addition, we examine in particular an algorithm based on transformation of beta variates and provide three useful guidelines so as to reduce its computer generation time. Simulation results show that the proposed algorithm significantly outperforms other approaches in terms of computer generation time, except in cases when all (or most) shape parameters are close to zero.
Profile monitoring has received increasingly attention in a wide range of applications in statist... more Profile monitoring has received increasingly attention in a wide range of applications in statistical process control (SPC). In this work, we propose a framework for monitoring nonparametric profiles in multi-dimensional data spaces. The framework has the following important features: (i) a flexible and computationally efficient smoothing technique, called Support Vector Regression, is employed to describe the relationship between the response variable and the explanatory variables; (ii) the usual structural assumptions on the residuals are not required; and (iii) the dependence structure for the within-profile observations is appropriately accommodated. Finally, real AIDS data collected from hospitals in Taiwan are used to illustrate and evaluate our proposed framework.
We consider a general model framework for acyclic stochastic processing networks with shared reso... more We consider a general model framework for acyclic stochastic processing networks with shared resources that has many applications in telecommunication, computer, and manufacturing systems. A dynamic control policy that utilizes the maximal matching (for scheduling) and the join-the-shortest-queue (for routing) discipline, is shown to maximize the throughput and stabilize the system in a sense called "uniform mean recurrence time property" under fairly mild stochastic assumptions. Owing to the non-Markovian nature of the states, system stability is established using a perturbed Lyapunov function method.
Shepp's urn model is a useful tool for analyzing the stopping-rule problems in economics and fina... more Shepp's urn model is a useful tool for analyzing the stopping-rule problems in economics and finance. In [R.W. Chen, A. Zame, C.T. Lin, H. Wu, A random version of Shepp's urn scheme, SIAM J. Discrete Math. 19 , Chen et al. considered a random version of Shepp's urn scheme and showed that a simple drawing policy (called "the k in the hole policy") can asymptotically maximize the expected value of the game. By extending the work done by Chen et al., this note considers a more general urn scheme that is better suited to real-life price models in which the short-term value might not fluctuate. Further, "the k in the hole policy" is shown to be asymptotically optimal for this new urn scheme.
Nonparametric profile monitoring B-spline Block bootstrap Confidence band Curve depth a b s t r a... more Nonparametric profile monitoring B-spline Block bootstrap Confidence band Curve depth a b s t r a c t Control charts have been widely used for monitoring the functional relationship between a response variable and some explanatory variable(s) (called profile) in various industrial applications. In this article, we propose an easy-to-implement framework for monitoring nonparametric profiles in both Phase I and Phase II of a control chart scheme. The proposed framework includes the following steps: (i) data cleaning; (ii) fitting B-spline models; (iii) resampling for dependent data using block bootstrap method; (iv) constructing the confidence band based on bootstrap curve depths; and (v) monitoring profiles online based on curve matching. It should be noted that, the proposed method does not require any structural assumptions on the data and, it can appropriately accommodate the dependence structure of the withinprofile observations. We illustrate and evaluate our proposed framework by using a real data set.
A software package, rBeta2009, developed to generate beta random numbers and Dirichlet random vec... more A software package, rBeta2009, developed to generate beta random numbers and Dirichlet random vectors in R is presented. The package incorporates state-of-the-art algorithms so as to minimize the computer generation time. In addition, it is designed in a way that (i) the generation efficiency is robust to changes of computer architecture; (ii) memory allocation is flexible; and (iii) the exported objects can be easily integrated with other software. The usage of this package is then illustrated and evaluated in terms of various performance metrics.
The power of uniform design (UD) has received great attention in the area of computer experiments... more The power of uniform design (UD) has received great attention in the area of computer experiments over the last two decades. However, when conducting a typical computer experiment, one finds many non-rectangular types of input domains on which traditional UD methods cannot be adequately applied. In this study, we propose a new UD method that is suitable for any type of design area. For practical implementation, we develop an efficient algorithm to construct a so-called nearly uniform design (NUD) and show that it approximates very well the UD solution for small sizes of experiment. By utilizing the proposed UD method, we also develop a methodology for estimating the target region of computer experiments. The methodology is sequential and aims to (i) provide adaptive models that predict well the output measures related to the experimental target; and (ii) minimize the number of experimental trials. Finally, we illustrate the developed methodology on various examples and show that, given the same experimental budget, it outperforms other approaches in estimating the prespecified target region of computer experiments.
The utilization of multiple fidelity simulators for the design and analysis of computer experimen... more The utilization of multiple fidelity simulators for the design and analysis of computer experiments has received increased attention in recent years. In this paper, we study the contour estimation problem for complex systems by considering two fidelity simulators. Our goal is to design a methodology of choosing the best suited simulator and input location for each simulation trial so that the overall estimation of the desired contour can be as good as possible under limited simulation resources. The proposed methodology is sequential and based on the construction of Gaussian process surrogate for the output measure of interest. We illustrate the methodology on a canonical queueing system and evaluate its efficiency via a simulation study.
ABSTRACT The validation of causal relationship between two groups of multivariate time series dat... more ABSTRACT The validation of causal relationship between two groups of multivariate time series data often requires the precedence knowledge of all variables. However, in practice one finds that some variables may be negligible in describing the underlying causal structure. In this article we provide an explicit definition of “non-informative variables” in a two-group causal relationship and introduce various automatic computer-search algorithms that can be utilized to extract informative variables based on a hypothesis testing procedure. The result allows us to represent a simplified causal relationship by using minimum possible information on two groups of variables.
Proceedings of the 2003 International Conference on Machine Learning and Cybernetics (IEEE Cat. No.03EX693), 2000
Simulation can provide insight to the behavior of a complex queueing system by identifying the re... more Simulation can provide insight to the behavior of a complex queueing system by identifying the response surface of several performance measures such as delays and backlogs. However, simulations of large systems are expensive both in terms of CPU time and use of available resources (e.g. processors). Thus, it is of paramount importance to carefully select the inputs of simulation in order to adequately capture the underlying response surface of interest and at the same time minimize the required number of simulation runs. In this study, we present a methodological framework for designing efficient simulations for complex networks. Our approach works in sequential and combines the methods of CART (Classification And Regression Trees) and the design of experiments. A generalized switch model is used to illustrate the proposed methodology and some useful applications are described.
2006 11th Intenational Workshop on Computer-Aided Modeling, Analysis and Design of Communication Links and Networks, 2000
Switched Processing Systems (SPS) capture the essence of a fundamental resource allocation proble... more Switched Processing Systems (SPS) capture the essence of a fundamental resource allocation problem in many modern communication, computer and manufacturing systems involving heterogeneous processors and multiple classes of job traffic flows. Recently, increased attention has been paid to the issue of improving quality of service (QoS) performance in terms of delays and backlogs of the associated scheduling policies, rather than simply maximizing the system's throughput. In this study, we investigate a class of throughput maximizing scheduling policies called MaxProduct policies. The objective is that through a use of dynamically changing "optimal" queue weights, the corresponding QoS performance measure -e.g. the average system delay-can be significantly improved. The proposed approach involves a statistical smoothing technique for tracking the system's workload and utilizes the result of how the MaxProduct policies drain out an initially placed workload in the shortest possible time. It is further shown that the proposed modification of the MaxProduct policy, achieves maximum throughput without requiring knowledge of the incoming traffic's statistics. The scheduling policy is illustrated on a small SPS subject to different types of input traffic.
We provide a simple proof of Wald's second identity for a class of problems that can be formulate... more We provide a simple proof of Wald's second identity for a class of problems that can be formulated as a fair coin-tossing game. The identity provides a use-ful technique for computing the expected stopping time and allows us to explore the second-order moment of the so-called heads-minus-tails distribution. We also introduce some interesting applications related to this simple identity.
There is increased interest in deploying charging station infrastructure for electric vehicles, d... more There is increased interest in deploying charging station infrastructure for electric vehicles, due to the increasing adoption of such vehicles to reduce emissions. However, there are a number of key challenges for providing high quality of service to such vehicles, stemming from technological reasons. One of them is due to the relative slow charging times and the other is due to the relative limited battery range. Hence, developing efficient routing strategies of electric vehicles requesting charging to stations that have available charging resources is an important component of the infrastructure. In this work, we propose a queueing modeling framework for the problem at hand and develop such routing strategies that optimize a performance metric related to vehicles' sojourn time in the system. By incorporating appropriate weights into the well-known dynamic routing discipline "Join-the-Shortest-Queue", we show that the proposed routing strategies not only do they maximize the queueing system's throughput, but also significantly mitigate the vehicle's sojourn time. The strategies are also adaptive in nature and responsive to changes in the speed of charging at the stations, the distribution of the vehicles' point of origin when requesting service, the traffic congestion level and the vehicle speed; all the above are novel aspects and compatible with the requirements of a modern electric vehicle charging infrastructure.
2007 IEEE International Conference on Communications, 2007
Switched Processing Systems (SPS) represent a canonical model for many areas of applications of c... more Switched Processing Systems (SPS) represent a canonical model for many areas of applications of communication, computer and manufacturing systems. They are characterized by flexible, interdependent service capabilities and multiple classes of job traffic flows. Recently, increased attention has been paid to the issue of improving quality of service (QoS) performance in terms of delays and backlogs of the associated scheduling policies, rather than simply maximizing the system's throughput. In this study, we investigate a measurement based dynamic service allocation policy that significantly improves performance with respect to delay metrics. The proposed policy solves a linear program at selected points in time that are in turn determined by a monitoring strategy that detects 'significant' changes in the intensities of the input processes. The proposed strategy is illustrated on a small SPS subject to different types of input traffic.
Genetic/transcriptional regulatory interactions are shown to predict partial components of signal... more Genetic/transcriptional regulatory interactions are shown to predict partial components of signaling pathways, which have been recognized as vital to complex human diseases. Both activator (A) and repressor (R) are known to coregulate their common target gene (T ). proposed to model this coregulation by a fixed second order response surface (called the RS algorithm), in which T is a function of A, R, and AR. Unfortunately, the RS algorithm did not result in a sufficient number of genetic interactions (GIs) when it was applied to a group of 51 yeast genes in a pilot study.Thus, we propose a data-driven second order model (DDSOM), an approximation to the non-linear transcriptional interactions, to infer genetic and transcriptional regulatory interactions. For each triplet of genes of interest (A, R, andT ), we regress the expression ofT at time t + 1 on the expression of A, R, and AR at time t. Next, these well-fitted regression models (viewed as points in R 3 ) are collected, and the center of these points is used to identify triples of genes having the A-R-T relationship or GIs. The DDSOM and RS algorithms are first compared on inferring transcriptional compensation interactions of a group of yeast genes in DNA synthesis and DNA repair using microarray gene expression data; the DDSOM algorithm results in higher modified true positive rate (about 75%) than that of the RS algorithm, checked against quantitative RT-polymerase chain reaction results.These validated GIs are reported, among which some coincide with certain interactions in DNA repair and genome instability pathways in yeast. This suggests that the DDSOM algorithm has potential to predict pathway components. Further, both algorithms are applied to predict transcriptional regulatory interactions of 63 yeast genes. Checked against the known transcriptional regulatory interactions queried from TRANSFAC, the proposed also performs better than the RS algorithm.
The identification of causal effects between two groups of time series has been an important topi... more The identification of causal effects between two groups of time series has been an important topic in a wide range of applications such as economics, engineering, medicine, neuroscience, and biology. In this paper, a simplified causal relationship (called trimmed Granger causality) based on the context of Granger causality and vector autoregressive (VAR) model is introduced. The idea is to characterize a subset of "important variables" for both groups of time series so that the underlying causal structure can be presented based on minimum variable information. When the VAR model is specified, explicit solutions are provided for the identification of important variables. When the parameters of the VAR model are unknown, an efficient statistical hypothesis testing procedure is introduced to estimate the solution. An example representing the stock indices of different countries is used to illustrate the proposed methods. In addition, a simulation study shows that the proposed methods significantly outperform the Lasso-type methods in terms of the accuracy of characterizing the simplified causal relationship.
Switched Processing Systems (SPS) represent canonical models for many communication and computer ... more Switched Processing Systems (SPS) represent canonical models for many communication and computer systems. Over the years, much research has been devoted to developing the best scheduling policies to optimize the various performance metrics of interest. These policies have mostly originated from the well-known MaxWeight discipline, which at any point in time switches the system into the service mode possessing "maximal matching" with the system state (e.g., queue-length, workload, etc.). However, for simplicity it is often assumed that the switching times between service modes are "negligible"-but this proves to be impractical in some applications. In this study, we propose a new scheduling strategy (called the Dynamic Cone Policy) for SPS, which includes substantial service-mode switching times. The goal is to maximize throughput and maintain system stability under fairly mild stochastic assumptions. For practical purposes, an extended scheduling strategy (called the Practical Dynamic Cone Policy) is developed to reduce the computational complexity of the Dynamic Cone Policy and at the same time mitigate job delay. A simulation study shows that the proposed practical policy clearly outperforms another throughput-maximizing policy called BatchAdapt, both in terms of the average and the 95th percentile of job delay for various types of input traffic.
Simulation can provide insight to the behavior of a complex queueing system by identifying the re... more Simulation can provide insight to the behavior of a complex queueing system by identifying the response surface of several performance measures such as delays and backlogs. However, simulations of large systems are expensive both in terms of CPU time ...
Journal of Statistical Planning and Inference, 2012
We consider the bandit problem with an infinite number of Bernoulli arms, of which the unknown pa... more We consider the bandit problem with an infinite number of Bernoulli arms, of which the unknown parameters are assumed to be i.i.d. random variables with a common distribution F. Our goal is to construct optimal strategies of choosing ''arms'' so that the expected long-run failure rate is minimized. We first review a class of strategies and establish their asymptotic properties when F is known. Based on the results, we propose a new strategy and prove that it is asymptotically optimal when F is unknown. Finally, we show that the proposed strategy performs well for a number of simulation scenarios.
Journal of Statistical Computation and Simulation, 2011
ABSTRACT In this article, we describe various well-known Dirichlet generation algorithms and eval... more ABSTRACT In this article, we describe various well-known Dirichlet generation algorithms and evaluate their performance in terms of the following criteria: (i) computer generation time, (ii) sensitivity, and (iii) goodness of fit. In addition, we examine in particular an algorithm based on transformation of beta variates and provide three useful guidelines so as to reduce its computer generation time. Simulation results show that the proposed algorithm significantly outperforms other approaches in terms of computer generation time, except in cases when all (or most) shape parameters are close to zero.
Profile monitoring has received increasingly attention in a wide range of applications in statist... more Profile monitoring has received increasingly attention in a wide range of applications in statistical process control (SPC). In this work, we propose a framework for monitoring nonparametric profiles in multi-dimensional data spaces. The framework has the following important features: (i) a flexible and computationally efficient smoothing technique, called Support Vector Regression, is employed to describe the relationship between the response variable and the explanatory variables; (ii) the usual structural assumptions on the residuals are not required; and (iii) the dependence structure for the within-profile observations is appropriately accommodated. Finally, real AIDS data collected from hospitals in Taiwan are used to illustrate and evaluate our proposed framework.
We consider a general model framework for acyclic stochastic processing networks with shared reso... more We consider a general model framework for acyclic stochastic processing networks with shared resources that has many applications in telecommunication, computer, and manufacturing systems. A dynamic control policy that utilizes the maximal matching (for scheduling) and the join-the-shortest-queue (for routing) discipline, is shown to maximize the throughput and stabilize the system in a sense called "uniform mean recurrence time property" under fairly mild stochastic assumptions. Owing to the non-Markovian nature of the states, system stability is established using a perturbed Lyapunov function method.
Shepp's urn model is a useful tool for analyzing the stopping-rule problems in economics and fina... more Shepp's urn model is a useful tool for analyzing the stopping-rule problems in economics and finance. In [R.W. Chen, A. Zame, C.T. Lin, H. Wu, A random version of Shepp's urn scheme, SIAM J. Discrete Math. 19 , Chen et al. considered a random version of Shepp's urn scheme and showed that a simple drawing policy (called "the k in the hole policy") can asymptotically maximize the expected value of the game. By extending the work done by Chen et al., this note considers a more general urn scheme that is better suited to real-life price models in which the short-term value might not fluctuate. Further, "the k in the hole policy" is shown to be asymptotically optimal for this new urn scheme.
Nonparametric profile monitoring B-spline Block bootstrap Confidence band Curve depth a b s t r a... more Nonparametric profile monitoring B-spline Block bootstrap Confidence band Curve depth a b s t r a c t Control charts have been widely used for monitoring the functional relationship between a response variable and some explanatory variable(s) (called profile) in various industrial applications. In this article, we propose an easy-to-implement framework for monitoring nonparametric profiles in both Phase I and Phase II of a control chart scheme. The proposed framework includes the following steps: (i) data cleaning; (ii) fitting B-spline models; (iii) resampling for dependent data using block bootstrap method; (iv) constructing the confidence band based on bootstrap curve depths; and (v) monitoring profiles online based on curve matching. It should be noted that, the proposed method does not require any structural assumptions on the data and, it can appropriately accommodate the dependence structure of the withinprofile observations. We illustrate and evaluate our proposed framework by using a real data set.
A software package, rBeta2009, developed to generate beta random numbers and Dirichlet random vec... more A software package, rBeta2009, developed to generate beta random numbers and Dirichlet random vectors in R is presented. The package incorporates state-of-the-art algorithms so as to minimize the computer generation time. In addition, it is designed in a way that (i) the generation efficiency is robust to changes of computer architecture; (ii) memory allocation is flexible; and (iii) the exported objects can be easily integrated with other software. The usage of this package is then illustrated and evaluated in terms of various performance metrics.
The power of uniform design (UD) has received great attention in the area of computer experiments... more The power of uniform design (UD) has received great attention in the area of computer experiments over the last two decades. However, when conducting a typical computer experiment, one finds many non-rectangular types of input domains on which traditional UD methods cannot be adequately applied. In this study, we propose a new UD method that is suitable for any type of design area. For practical implementation, we develop an efficient algorithm to construct a so-called nearly uniform design (NUD) and show that it approximates very well the UD solution for small sizes of experiment. By utilizing the proposed UD method, we also develop a methodology for estimating the target region of computer experiments. The methodology is sequential and aims to (i) provide adaptive models that predict well the output measures related to the experimental target; and (ii) minimize the number of experimental trials. Finally, we illustrate the developed methodology on various examples and show that, given the same experimental budget, it outperforms other approaches in estimating the prespecified target region of computer experiments.
The utilization of multiple fidelity simulators for the design and analysis of computer experimen... more The utilization of multiple fidelity simulators for the design and analysis of computer experiments has received increased attention in recent years. In this paper, we study the contour estimation problem for complex systems by considering two fidelity simulators. Our goal is to design a methodology of choosing the best suited simulator and input location for each simulation trial so that the overall estimation of the desired contour can be as good as possible under limited simulation resources. The proposed methodology is sequential and based on the construction of Gaussian process surrogate for the output measure of interest. We illustrate the methodology on a canonical queueing system and evaluate its efficiency via a simulation study.
ABSTRACT The validation of causal relationship between two groups of multivariate time series dat... more ABSTRACT The validation of causal relationship between two groups of multivariate time series data often requires the precedence knowledge of all variables. However, in practice one finds that some variables may be negligible in describing the underlying causal structure. In this article we provide an explicit definition of “non-informative variables” in a two-group causal relationship and introduce various automatic computer-search algorithms that can be utilized to extract informative variables based on a hypothesis testing procedure. The result allows us to represent a simplified causal relationship by using minimum possible information on two groups of variables.
Uploads
Papers by Ying-Chao Hung