I. Lagaris
University of Ioannina/Greece, Computer Science & Engineering, Faculty Member
Research Interests:
ABSTRACT The appearance of bound states with large binding energies of several hundred MeV in the three-body system, known as bound state collapse, is investigated. For this purpose three classes of two-body potentials are employed; local... more
ABSTRACT The appearance of bound states with large binding energies of several hundred MeV in the three-body system, known as bound state collapse, is investigated. For this purpose three classes of two-body potentials are employed; local potentials equivalent to nonlocal interactions possessing a continuum bound state, in addition to the usual negative-energy bound state; local potentials with a strong attractive well sustaining a forbidden state; and supersymmetric transformation potentials. It is first shown that local potentials equivalent to the above nonlocal ones have a strong attractive well in the interior region which supports, in addition to the physical deuteron state, a second bound state (usually called a pseudobound state) with a large binding energy, which is responsible for the bound state collapse in the three-body (and in general to the N-body) system. Second, it is shown that local potentials with a forbidden state also generate a three-body bound state collapse, implying that the role played by the forbidden state is similar to the one played by the pseudobound state. Finally, it is shown that the removal of the forbidden state via supersymmetric transformations also results in the disappearance of the collapse. Thus one can safely argue that the presence of unphysical bound states with large binding energies in the two-body system is responsible for the bound state collapse in the three-body system.
Research Interests:
Research Interests:
Research Interests:
Research Interests: Engineering, Computer Science, Algorithms, Artificial Intelligence, Fluorescence, and 23 moreFluorescence Microscopy, Image Analysis, Molecular Biophysics, Gene expression, In Situ Hybridization, Parameter estimation, DNA, Gaussian processes, Humans, Computer Simulation, Maximum Likelihood, Animals, Cross Validation, Maximum a Posteriori, Maximum Likelihood Estimation, Gaussian Mixture Model, Gaussian mixture Models, Mixture Model, Software Tool, Expectation Maximization (EM) Algorithm, Gene expression profiling, Maximum Likelihood Method, and Markov Random Field
Research Interests:
Abstract. A new training algorithm is presented for delayed reinforcement learning problems,that does not assume,the existence of a critic model and employs the polytope optimization algorithm to adjust the weights of the action network... more
Abstract. A new training algorithm is presented for delayed reinforcement learning problems,that does not assume,the existence of a critic model and employs the polytope optimization algorithm to adjust the weights of the action network so that a simple direct measure,of the training performance is maximized. Experimental results from the application of the method,to the pole balancing prob- lem indicate improved,training performance,compared,with critic-based and genetic reinforcement approaches. Key words: reinforcement learning, neurocontrol, optimization, polytope algorithm, pole balancing,
Research Interests:
Gaussian mixture models (GMMs) constitute a well-known type of probabilistic neural networks. One of their many successful applications is in image segmentation, where spatially constrained mixture models have been trained using the... more
Gaussian mixture models (GMMs) constitute a well-known type of probabilistic neural networks. One of their many successful applications is in image segmentation, where spatially constrained mixture models have been trained using the expectation-maximization (EM) framework. In this letter, we elaborate on this method and propose a new methodology for the M-step of the EM algorithm that is based on a