CiteSeerX — Citation Query The elements of statistical learning (2nd edition

来自 : 发布时间：2024-05-06

\"... Abstract—Hair segmentation is challenging due to the diverse appearance, irregular region boundary and the influence of complex background. To deal with this problem, we propose a novel method, named Isomorphic Manifold Inference (IMI). Given a head-shoulder image, a Coarse Hair Probability Map (Coa ...\" Abstract Cited by 1 (0 self) - Add to MetaCart Abstract—Hair segmentation is challenging due to the diverse appearance, irregular region boundary and the influence of complex background. To deal with this problem, we propose a novel method, named Isomorphic Manifold Inference (IMI). Given a head-shoulder image, a Coarse Hair Probability Map (Coarse HPM), each element of which represents the probability of the pixel being hair, is initially calculated by exploring hair location and color priors. Then, based on an observation that similar Coarse HPMs imply similar segmentations, we formulate Coarse HPM and corresponding ground segmentation (Optimal HPM) as a pair of isomorphic manifolds. Under this formulation, final hair segmentation is inferred from the Coarse HPM with regression techniques. In this way, the IMI implicitly exploits the hair-specific prior embodied in the training set. Extensive experimental comparisons are conducted and the results strongly encourage the method. The generality of IMI to other class-specific image segmentation is also discussed. Keywords-Hair segmentation; Shape prior; Isomorphic

Citation Context

...r distribution.sUnder this isomorphic manifold formulation, we proposesIsomorphic Manifold Inference (IMI) method to compute asmore accurate HPM (Refined HPM), by employing somesregression techniques =-=[25]-=-. In some sense, our IMI is a ‘noisefiltering’ procedure, which exploits the hair prior implied insthe training set to refine the Coarse HPMs. In IMI, appropriatesconstraints are imposed on regression...

Arlene S. Ash, Phd Stephen, E. Fienberg, Phd Thomas, A. Louis, Sharon-lise T. Norm, Phd Thérèse, A. Stukel, Phd Jessica Utts \"... is supporting a committee appointed by the Committee of Presidents of Statistical Societies (COPSS) to address statistical issues identified by the CMS and stakeholders about CMS’s approach to modeling hospital quality based on outcomes. In the spring of 2011, with the direct support of YN-HHSC/CORE ...\" Abstract Cited by 1 (0 self) - Add to MetaCart is supporting a committee appointed by the Committee of Presidents of Statistical Societies (COPSS) to address statistical issues identified by the CMS and stakeholders about CMS’s approach to modeling hospital quality based on outcomes. In the spring of 2011, with the direct support of YN-HHSC/CORE, COPSS formed a committee comprised of one member from each of its constituent societies, a chair, and a staff member from the American Statistical Association, and held a preliminary meeting in April. In June, YNHHSC/CORE executed a subcontract with COPSS under its CMS contract to support the development of a White Paper on statistical modeling. Specifically, YNHHSC/CORE contracted with COPSS to \"provide guidance on statistical approaches...when estimating performance metrics, ” and \"consider and discuss concerns commonly raised by stakeholders (hospitals, consumer, and insurers) about the use of \"hierarchical generalized linear models in profiling hospital quality. The committee convened in June and August of 2011, and exchanged a wide variety of materials. To ensure the committee’s independence, YNHHSC/CORE did not comment on the white paper findings, and CMS pre-cleared COPSS ’ publication of an academic manuscript based on the White Paper. The committee thanks COPSS and especially its chair, Xihong Lin of the Harvard School of Public Health; and staff of the American Statistical Association, especially Steve Pierson and Keith Crank, for their efforts in establishing the committee and coordinating its work. We thank Darcey Cobbs-

Citation Context

...s amongst patient-level predictors, use of generalized additive models or splines (Crainiceanu et al., 2007; Wood, 2006), classification trees, random forests and boosting (Berk, 2008; Breiman, 2001; =-=Hastie et al., 2009-=-; McCaffrey et al., 2004) and similar approaches. Model comparisons and evaluations include likelihood-based approaches such as AIC and BIC, and for hierarchical models DIC (Bayarri and Castellanos, 2...

Mai Hamdalla, David Grant, Ion M, Dennis Hill, Sanguthevar Rajasekaran, Reda Ammar - in 2012 IEEE 2nd International Conference on Computational Advances in Bio and medical Sciences (ICCABS), (NV \"... Abstract—Metabolomics is a rapidly growing field studying the small-molecule metabolite profile of a biological organism. Studying metabolism has a potential to contribute to biomedical research as well as drug discovery. One of the current challenges in metabolomics is the identification of unknown ...\" Abstract Cited by 1 (1 self) - Add to MetaCart Abstract—Metabolomics is a rapidly growing field studying the small-molecule metabolite profile of a biological organism. Studying metabolism has a potential to contribute to biomedical research as well as drug discovery. One of the current challenges in metabolomics is the identification of unknown metabolites as existing chemical databases are incomplete. We present a novel way of utilizing known mammalian metabolites in an effort to identify unknown ones. The system relies on a mammalian scaffolds database to aid the classification process. The results show that 96 % of the mammalian compounds were identified as truly mammalian in a leave-one-out experiment. The system was also tested with a random set of synthetic compounds, downloaded from ChemBridge and ChemSynthesis databases. The system was able to eliminate 54 % of the set, leaving 46 % of the compounds as potentially unknown mammalian metabolites.

Citation Context

...c compounds (retrieved from ChemBridge andsChemSynthesis databases).sCross Validation (CV) is one of the simplest and mostswidely used methods for estimating the accuracy ofsclassification algorithms =-=[22]-=-. Briefly, both the syntheticsand mammalian compounds were randomly split in half;sone half for training the model and the other half for testingsit. The training half was randomly split into K roughl...

\"... Abstract: Contemporary discussions of education in blended-learning environments increasingly emphasize the social nature of learning which emphasizes interactions among students, or among students and instructors. These interactions can occur asynchronously using a text based discussion forum. A te ...\" Abstract Cited by 1 (0 self) - Add to MetaCart Abstract: Contemporary discussions of education in blended-learning environments increasingly emphasize the social nature of learning which emphasizes interactions among students, or among students and instructors. These interactions can occur asynchronously using a text based discussion forum. A text-based discussion forum, however, may not work well for all participants as some find it difficult to explain complex concepts in words, while others complain of being misunderstood due to the absence of verbal cues. In this study, we investigated the use of a Wimba Voice Board to support asynchronous voice discussion. A quasi-experiment research design involving two classes of undergraduate students was conducted. One of the classes (n = 24 students) used the Wimba Voice Board while the other (n = 18 students) used a text discussion forum in BlackBoard. The results of an independent t-test analysis suggested that there was no significant difference in the students ’ degree of participation in the two classes, asynchronous voice discuss class (M = 2.92, SD = 1.586) and text discussion class (M = 2.78, SD = 1.353), (t = 0.299, df = 40, p = 0.767) at the 0.05 level of significance. However, the online discussion appeared to be more sustained in the asynchronous voice discussion group. Analyses of the students ’ reflection data suggested that asynchronous voice discussion have several advantages over text forums. Specifically, an asynchronous voice discussion:

Citation Context

...sof these data is approached in the spirit of \"training data,” or \"statistical learning,” in which earlysanalyses are used to improve the collection and analysis plans for subsequent data (Berk 2006;s=-=Hastie et al, 2009-=-).sWe have focussed initially on the measurement properties of the variables and onsplanning for analysis of data from upcoming semesters.sAnalyses of outcomes comparing control andsexperimental group...

\"... The ability to discover the effects of actions and apply this knowledge during goal-oriented action selection is a fundamental requirement of embodied intelligent agents. This requirement is most clearly demonstrated when the agent, or ...\" Abstract - Add to MetaCart The ability to discover the effects of actions and apply this knowledge during goal-oriented action selection is a fundamental requirement of embodied intelligent agents. This requirement is most clearly demonstrated when the agent, or The Copss-cms, Arlene S. Ash, Phd Stephen, E. Fienberg, Phd Thomas, A. Louis, Sharon-lise T. Norm, A. Stukel, Phd Jessica Utts \"... The Centers for Medicare and Medicaid Services (CMS), through a subcontract with Yale New ...\" Abstract - Add to MetaCart The Centers for Medicare and Medicaid Services (CMS), through a subcontract with Yale New

Citation Context

...s amongst patient-level predictors, usesof generalized additive models or splines (Crainiceanu et al., 2007; Wood, 2006), classificationstrees, random forests and boosting (Berk, 2008; Breiman, 2001; =-=Hastie et al., 2009-=-; McCaffreyset al., 2004) and similar approaches. Model comparisons and evaluations include likelihood-basedsapproaches such as AIC and BIC, and for hierarchical models DIC (Bayarri and Castellanos, 2...

\"... This chapter gives an overview of statistical methods used in high-energy physics. In statistics, we are interested in using a given sample of data to make inferences about a probabilistic model, e.g., to assess the model’s validity or to determine the values of its parameters. There are two main ap ...\" Abstract - Add to MetaCart This chapter gives an overview of statistical methods used in high-energy physics. In statistics, we are interested in using a given sample of data to make inferences about a probabilistic model, e.g., to assess the model’s validity or to determine the values of its parameters. There are two main approaches to statistical inference, which we may call frequentist and Bayesian. In frequentist statistics, probability is interpreted as the frequency of the outcome of a repeatable experiment. The most important tools in this framework are parameter estimation, covered in Section 32.1, and statistical tests, discussed in Section 32.2. Frequentist confidence intervals, which are constructed so as to cover the true value of a parameter with a specified probability, are treated in Section 32.3.2. Note that in frequentist statistics one does not define a probability for a hypothesis or for a parameter. Frequentist statistics provides the usual tools for reporting the outcome of an experiment objectively, without needing to incorporate prior beliefs concerning the parameter being measured or the theory being tested. As such, they are used for reporting most measurements and their statistical uncertainties in high-energy physics. \"... This chapter gives an overview of statistical methods used in high-energy physics. In statistics, we are interested in using a given sample of data to make inferences about a probabilistic model, e.g., to assess the model’s validity or to determine the values of its parameters. There are two main ap ...\" Abstract - Add to MetaCart This chapter gives an overview of statistical methods used in high-energy physics. In statistics, we are interested in using a given sample of data to make inferences about a probabilistic model, e.g., to assess the model’s validity or to determine the values of its parameters. There are two main approaches to statistical inference, which we may call frequentist and Bayesian. In frequentist statistics, probability is interpreted as the frequency of the outcome of a repeatable experiment. The most important tools in this framework are parameter estimation, covered in Section 32.1, and statistical tests, discussed in Section 32.2. Frequentist confidence intervals, which are constructed so as to cover the true value of a parameter with a specified probability, are treated in Section 32.3.2. Note that in frequentist statistics one does not define a probability for a hypothesis or for a parameter. Frequentist statistics provides the usual tools for reporting the outcome of an experiment objectively, without needing to incorporate prior beliefs concerning the parameter being measured or the theory being tested. As such, they are used for reporting most measurements and their statistical uncertainties in high-energy physics. Chun-nan Hsu, Cheng-ju Kuo, Congxing Cai, Sarah A. Pendergrass, Marylyn D. Ritchie, Jose Luis Ambite \"... Accurate phenotype mapping will play an important role in facilitating Phenome-Wide Association Studies (PheWAS), and potentially in other phenomics based studies. The Phe-WAS approach investigates the association between genetic variation and an extensive range of phenotypes in a high-throughput ma ...\" Abstract - Add to MetaCart Accurate phenotype mapping will play an important role in facilitating Phenome-Wide Association Studies (PheWAS), and potentially in other phenomics based studies. The Phe-WAS approach investigates the association between genetic variation and an extensive range of phenotypes in a high-throughput manner to better understand the impact of genetic variations on multiple phenotypes. Herein we define the phenotype mapping problem posed by PheWAS analyses, discuss the challenges, and present a machine-learning solution. Our key ideas include the use of weighted Jaccard features and term augmentation by dictionary lookup. When compared to string similarity metric-based features, our approach improves the F-score from 0.59 to 0.73. With augmentation we show further improvement in F-score to 0.89. For terms not covered by the dictionary, we use transitive closure inference and reach an F-score of 0.91, close to a level sufficient for practical use. We also show that our model generalizes well to phenotypes not used in our training dataset. 1