Dr. Micheas's Personal Sketch and Research Interests

Athanasios Christou Micheas
Education: B.S. in Mathematics(1995), University of Ioannina, Department of Mathematics, Greece
Majors: Statistics, Computer Science
M.Sc. in Probability, Statistics and Operational Research(1997), University of Ioannina, Department of Mathematics, Section of Statistics, Greece
M.Sc. in Statistics(1998), University of Connecticut, Department of Statistics
Ph.D. in Statistics(2001), University of Connecticut, Department of Statistics
July, 2001, under the guidance of Professor Dipak K. Dey.
Selected Research Interests

Shape Analysis
The study of shapes that appear in experiments, has captured the attention of researchers in a variety of scientific contexts. From image analysis of Magnetic Resonance Images(MRI) to genetics, biology and a host of others, new methodologies are sorely needed in order to provide a better understanding of the shape of an object as well as help predict the shape of an object. Moreover, studying shape differences, either in average shape or shape variability, can provide medical doctors for example, with a new way of diagnosis and treatment of a certain disease, or paleontologists with a new method of categorizing specimens in certain age periods, and so forth.
Our goal is to create new methodologies that address these questions. We describe shapes by using "landmarks". Landmarks are points of correspondence on each shape, that match between and within populations of shapes. Then describing these landmarks stochastically, using some "shape distribution", allows us to infer about the average(modal) shape of a population of shapes, as well as assert shape differences between two or more populations of shapes. Bayesian as well as classical methods of estimation are being developed.

Bayesian Loss Robustness
The problem of robustness has always been an important element of statistics. In a Bayesian framework, there exists a vast literature that mainly concentrates on sensitivity analysis regarding choice of priors. There are only a few papers relating to loss robustness. The delay is probably due to the fact that robustness problems in the prior distribution are probably more important than loss robustness problems. Furthermore, when a statistical analysis is performed, there might be pressure of time or cost constraints that don't allow us to consider repeating the analysis under a different loss. Even so, a well performed analysis requires the examination of different loss functions in order to determine the loss that will allow the experimenter to make the optimum decision.
In this context we are interested in the development of new measures of loss robustness in a Bayesian framework, based usually on the Posterior Expected Loss (PEL), of the action we take. A variety of measures, including range of PEL, posterior regret, and range of MAP loss, can be investigated under different classes of loss functions, like LINEX(Linear Exponential) and Hellinger loss functions. Application of these methods is desired for different models for the data, including the continuous exponential family and the discrete power series family of distributions.

Go to the department homepage.; Majors: Statistics, Computer Science
M.Sc. in Probability, Statistics and Operational Research(1997), University of Ioannina, Department of Mathematics, Section of Statistics, Greece
M.Sc. in Statistics(1998), University of Connecticut, Department of Statistics
Ph.D. in Statistics(2001), University of Connecticut, Department of Statistics
July, 2001, under the guidance of Professor Dipak K. Dey.
Selected Research Interests

Shape Analysis
The study of shapes that appear in experiments, has captured the attention of researchers in a variety of scientific contexts. From image analysis of Magnetic Resonance Images(MRI) to genetics, biology and a host of others, new methodologies are sorely needed in order to provide a better understanding of the shape of an object as well as help predict the shape of an object. Moreover, studying shape differences, either in average shape or shape variability, can provide medical doctors for example, with a new way of diagnosis and treatment of a certain disease, or paleontologists with a new method of categorizing specimens in certain age periods, and so forth.
Our goal is to create new methodologies that address these questions. We describe shapes by using "landmarks". Landmarks are points of correspondence on each shape, that match between and within populations of shapes. Then describing these landmarks stochastically, using some "shape distribution", allows us to infer about the average(modal) shape of a population of shapes, as well as assert shape differences between two or more populations of shapes. Bayesian as well as classical methods of estimation are being developed.

Bayesian Loss Robustness
The problem of robustness has always been an important element of statistics. In a Bayesian framework, there exists a vast literature that mainly concentrates on sensitivity analysis regarding choice of priors. There are only a few papers relating to loss robustness. The delay is probably due to the fact that robustness problems in the prior distribution are probably more important than loss robustness problems. Furthermore, when a statistical analysis is performed, there might be pressure of time or cost constraints that don't allow us to consider repeating the analysis under a different loss. Even so, a well performed analysis requires the examination of different loss functions in order to determine the loss that will allow the experimenter to make the optimum decision.
In this context we are interested in the development of new measures of loss robustness in a Bayesian framework, based usually on the Posterior Expected Loss (PEL), of the action we take. A variety of measures, including range of PEL, posterior regret, and range of MAP loss, can be investigated under different classes of loss functions, like LINEX(Linear Exponential) and Hellinger loss functions. Application of these methods is desired for different models for the data, including the continuous exponential family and the discrete power series family of distributions.

Go to the department homepage.; July, 2001, under the guidance of Professor Dipak K. Dey.
Selected Research Interests: Shape Analysis
The study of shapes that appear in experiments, has captured the attention of researchers in a variety of scientific contexts. From image analysis of Magnetic Resonance Images(MRI) to genetics, biology and a host of others, new methodologies are sorely needed in order to provide a better understanding of the shape of an object as well as help predict the shape of an object. Moreover, studying shape differences, either in average shape or shape variability, can provide medical doctors for example, with a new way of diagnosis and treatment of a certain disease, or paleontologists with a new method of categorizing specimens in certain age periods, and so forth.
Our goal is to create new methodologies that address these questions. We describe shapes by using "landmarks". Landmarks are points of correspondence on each shape, that match between and within populations of shapes. Then describing these landmarks stochastically, using some "shape distribution", allows us to infer about the average(modal) shape of a population of shapes, as well as assert shape differences between two or more populations of shapes. Bayesian as well as classical methods of estimation are being developed.

Bayesian Loss Robustness
The problem of robustness has always been an important element of statistics. In a Bayesian framework, there exists a vast literature that mainly concentrates on sensitivity analysis regarding choice of priors. There are only a few papers relating to loss robustness. The delay is probably due to the fact that robustness problems in the prior distribution are probably more important than loss robustness problems. Furthermore, when a statistical analysis is performed, there might be pressure of time or cost constraints that don't allow us to consider repeating the analysis under a different loss. Even so, a well performed analysis requires the examination of different loss functions in order to determine the loss that will allow the experimenter to make the optimum decision.
In this context we are interested in the development of new measures of loss robustness in a Bayesian framework, based usually on the Posterior Expected Loss (PEL), of the action we take. A variety of measures, including range of PEL, posterior regret, and range of MAP loss, can be investigated under different classes of loss functions, like LINEX(Linear Exponential) and Hellinger loss functions. Application of these methods is desired for different models for the data, including the continuous exponential family and the discrete power series family of distributions.

Go to the department homepage.; The study of shapes that appear in experiments, has captured the attention of researchers in a variety of scientific contexts. From image analysis of Magnetic Resonance Images(MRI) to genetics, biology and a host of others, new methodologies are sorely needed in order to provide a better understanding of the shape of an object as well as help predict the shape of an object. Moreover, studying shape differences, either in average shape or shape variability, can provide medical doctors for example, with a new way of diagnosis and treatment of a certain disease, or paleontologists with a new method of categorizing specimens in certain age periods, and so forth.; Our goal is to create new methodologies that address these questions. We describe shapes by using "landmarks". Landmarks are points of correspondence on each shape, that match between and within populations of shapes. Then describing these landmarks stochastically, using some "shape distribution", allows us to infer about the average(modal) shape of a population of shapes, as well as assert shape differences between two or more populations of shapes. Bayesian as well as classical methods of estimation are being developed.

Bayesian Loss Robustness
The problem of robustness has always been an important element of statistics. In a Bayesian framework, there exists a vast literature that mainly concentrates on sensitivity analysis regarding choice of priors. There are only a few papers relating to loss robustness. The delay is probably due to the fact that robustness problems in the prior distribution are probably more important than loss robustness problems. Furthermore, when a statistical analysis is performed, there might be pressure of time or cost constraints that don't allow us to consider repeating the analysis under a different loss. Even so, a well performed analysis requires the examination of different loss functions in order to determine the loss that will allow the experimenter to make the optimum decision.
In this context we are interested in the development of new measures of loss robustness in a Bayesian framework, based usually on the Posterior Expected Loss (PEL), of the action we take. A variety of measures, including range of PEL, posterior regret, and range of MAP loss, can be investigated under different classes of loss functions, like LINEX(Linear Exponential) and Hellinger loss functions. Application of these methods is desired for different models for the data, including the continuous exponential family and the discrete power series family of distributions.

Go to the department homepage.; The problem of robustness has always been an important element of statistics. In a Bayesian framework, there exists a vast literature that mainly concentrates on sensitivity analysis regarding choice of priors. There are only a few papers relating to loss robustness. The delay is probably due to the fact that robustness problems in the prior distribution are probably more important than loss robustness problems. Furthermore, when a statistical analysis is performed, there might be pressure of time or cost constraints that don't allow us to consider repeating the analysis under a different loss. Even so, a well performed analysis requires the examination of different loss functions in order to determine the loss that will allow the experimenter to make the optimum decision.; In this context we are interested in the development of new measures of loss robustness in a Bayesian framework, based usually on the Posterior Expected Loss (PEL), of the action we take. A variety of measures, including range of PEL, posterior regret, and range of MAP loss, can be investigated under different classes of loss functions, like LINEX(Linear Exponential) and Hellinger loss functions. Application of these methods is desired for different models for the data, including the continuous exponential family and the discrete power series family of distributions.
Go to the department homepage.

Athanasios Christou Micheas

Education

Selected Research Interests