The Experts below are selected from a list of 810906 Experts worldwide ranked by ideXlab platform
Michael C. Ferris - One of the best experts on this subject based on the ideXlab platform.
-
Design optimization of a robust sleeve antenna for hepatic microwave ablation.
Physics in medicine and biology, 2008Co-Authors: Punit Prakash, Geng Deng, M.c. Converse, John G. Webster, David M. Mahvi, Michael C. FerrisAbstract:We describe the application of a Bayesian variable-Number Sample-path (VNSP) optimization algorithm to yield a robust design for a floating sleeve antenna for hepatic microwave ablation. Finite element models are used to generate the electromagnetic (EM) field and thermal distribution in liver given a particular design. Dielectric properties of the tissue are assumed to vary within ± 10% of average properties to simulate the variation among individuals. The Bayesian VNSP algorithm yields an optimal design that is a 14.3% improvement over the original design and is more robust in terms of lesion size, shape and efficiency. Moreover, the Bayesian VNSP algorithm finds an optimal solution saving 68.2% simulation of the evaluations compared to the standard Sample-path optimization method.
-
Variable-Number Sample-Path Optimization
Mathematical Programming, 2007Co-Authors: Geng Deng, Michael C. FerrisAbstract:The Sample-path method is one of the most important tools in simulation-based optimization. The basic idea of the method is to approximate the expected simulation output by the average of Sample observations with a common random Number sequence. In this paper, we describe a new variant of Powell’s unconstrained optimization by quadratic approximation (UOBYQA) method, which integrates a Bayesian variable-Number Sample-path (VNSP) scheme to choose appropriate Number of Samples at each iteration. The statistically accurate scheme determines the Number of simulation runs, and guarantees the global convergence of the algorithm. The VNSP scheme saves a significant amount of simulation operations compared to general purpose ‘fixed-Number’ Sample-path methods. We present numerical results based on the new algorithm.
Ping Guo - One of the best experts on this subject based on the ideXlab platform.
-
A study of regularized Gaussian classifier in high-dimension small Sample set case based on MDL principle with application to spectrum recognition
Pattern Recognition, 2008Co-Authors: Ping Guo, Yunde Jia, Michael R. LyuAbstract:In classifying high-dimensional patterns such as stellar spectra by a Gaussian classifier, the covariance matrix estimated with a small-Number Sample set becomes unstable, leading to degraded classification accuracy. In this paper, we investigate the covariance matrix estimation problem for small-Number Samples with high dimension setting based on minimum description length (MDL) principle. A new covariance matrix estimator is developed, and a formula for fast estimation of regularization parameters is derived. Experiments on spectrum pattern recognition are conducted to investigate the classification accuracy with the developed covariance matrix estimator. Higher classification accuracy results are obtained and demonstrated in our new approach.
-
IJCNN - On the study of BKYY cluster Number selection criterion for small Sample data set with bootstrap technique
IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339), 1Co-Authors: Ping GuoAbstract:The Bayesian-Kullback ying-yang (BKYY) learning theory and system has been proposed by Xu (1995, 1997), and one special case of ying-yang system can provide the model selection criteria for selecting the Number of clusters in the clustering analysis. In this paper, we present an experimental study of this cluster Number selection criterion in a small Number Sample set case. The results show that the criterion performed reasonable well when mixture parameters were estimated by incorporating a bootstrap technique with the EM algorithm.
R. Tran Manh Sung - One of the best experts on this subject based on the ideXlab platform.
-
Influence of the quantity of nonspecific DNA and repeated freezing and thawing of Samples on the quantification of DNA by the Light Cycler.
Journal of microbiological methods, 2003Co-Authors: B. Bellete, P. Flori, J. Hafid, H. Raberin, R. Tran Manh SungAbstract:Abstract Quantification of DNA in real-time using the Light Cycler® is increasingly being used for the detection and follow-up of various infectious and other diseases. We evaluated the effect of two parameters, namely the presence of nonspecific DNA and prior repeated freezing and thawing on the accurate quantification of DNA extracts from the RH strain of Toxoplasma gondii by the SYBR Green I and the Hybridization Probe techniques. For both parameters, a high copy Number Sample containing 5×10 5 parasites/extract and a low copy Number Sample containing 100 parasites/extract were tested. Reliable quantification was possible in the presence of up to 200 ng of nonspecific DNA by the SYBR Green I technique and up to 1000 ng by the Hybridization Probe technique as compared to the company threshold of 50 and 500 ng, respectively. As tissue Samples usually contain more than 200 ng of nonspecific DNA, the ideal choice is the Hybridization Probe technique. The stability of DNA extracts after repeated freeze–thaw cycles was found to be dependent on the volume in which they were stored. Samples stored in 100-μl total volumes were not stable after 3 freeze–thaw cycles, whereas those stored in 1-ml total volumes were stable after 14 freeze–thaw cycles.
Michael R. Lyu - One of the best experts on this subject based on the ideXlab platform.
-
A study of regularized Gaussian classifier in high-dimension small Sample set case based on MDL principle with application to spectrum recognition
Pattern Recognition, 2008Co-Authors: Ping Guo, Yunde Jia, Michael R. LyuAbstract:In classifying high-dimensional patterns such as stellar spectra by a Gaussian classifier, the covariance matrix estimated with a small-Number Sample set becomes unstable, leading to degraded classification accuracy. In this paper, we investigate the covariance matrix estimation problem for small-Number Samples with high dimension setting based on minimum description length (MDL) principle. A new covariance matrix estimator is developed, and a formula for fast estimation of regularization parameters is derived. Experiments on spectrum pattern recognition are conducted to investigate the classification accuracy with the developed covariance matrix estimator. Higher classification accuracy results are obtained and demonstrated in our new approach.
Geng Deng - One of the best experts on this subject based on the ideXlab platform.
-
Design optimization of a robust sleeve antenna for hepatic microwave ablation.
Physics in medicine and biology, 2008Co-Authors: Punit Prakash, Geng Deng, M.c. Converse, John G. Webster, David M. Mahvi, Michael C. FerrisAbstract:We describe the application of a Bayesian variable-Number Sample-path (VNSP) optimization algorithm to yield a robust design for a floating sleeve antenna for hepatic microwave ablation. Finite element models are used to generate the electromagnetic (EM) field and thermal distribution in liver given a particular design. Dielectric properties of the tissue are assumed to vary within ± 10% of average properties to simulate the variation among individuals. The Bayesian VNSP algorithm yields an optimal design that is a 14.3% improvement over the original design and is more robust in terms of lesion size, shape and efficiency. Moreover, the Bayesian VNSP algorithm finds an optimal solution saving 68.2% simulation of the evaluations compared to the standard Sample-path optimization method.
-
Variable-Number Sample-Path Optimization
Mathematical Programming, 2007Co-Authors: Geng Deng, Michael C. FerrisAbstract:The Sample-path method is one of the most important tools in simulation-based optimization. The basic idea of the method is to approximate the expected simulation output by the average of Sample observations with a common random Number sequence. In this paper, we describe a new variant of Powell’s unconstrained optimization by quadratic approximation (UOBYQA) method, which integrates a Bayesian variable-Number Sample-path (VNSP) scheme to choose appropriate Number of Samples at each iteration. The statistically accurate scheme determines the Number of simulation runs, and guarantees the global convergence of the algorithm. The VNSP scheme saves a significant amount of simulation operations compared to general purpose ‘fixed-Number’ Sample-path methods. We present numerical results based on the new algorithm.
