Abordagens para aprendizado semissupervisionado multirrótulo e hierárquico

Metz, Jean

doi:10.11606/t.55.2011.tde-13012012-144607

Cited by 5 publications

(9 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…͑14b͒, the , , 1, is replaced, respectively, by xЈ,yЈ,zЈ defined in this paper. As in their method, the linear solution results in two rotation matrices and in a unit translation vector, t. The magnitude of the translation vector, ͉t͉, is given by the equation, 16 ͉t͉ϭ…”

Section: Methodsmentioning

confidence: 99%

Determination of three‐dimensional positions of known sparse objects from a single projection

Hoffmann

Esthappan

1997

Medical Physics

View full text Add to dashboard Cite

A new technique is developed for accurate determination of the three-dimensional position and orientation of known sparse objects, e.g., a configuration of points, from a single-perspective projection. In this technique, a computer model of the known object is translated and rotated so as to align it optimally in a least-squares sense with the projection lines connecting the image points with the focal spot by using a modification of the projection-Procrustes technique. The translational and rotational adjustments are repeated iteratively until the angular change between iterations is less than 0.25 degree. Simulations indicate that, for rms input image errors of 0.03 cm, the three-dimensional positions and orientations can be determined to within approximately 0.2 cm and 0.3 degree for a wide range of initially guessed positions and orientations, and positions can be determined with an accuracy of approximately 0.3 cm for objects having as few as four points. In phantom experiments, three-dimensional positions and orientations of a cube phantom were reproducibly determined to within 0.23 cm and 0.13 degree. The entire calculation requires only 10 s on a VAX 3500 to converge to the solution. The accuracy, precision, and speed of the technique indicate that it will be a useful tool for determination of three-dimensional positions and orientations of known sparse objects.

show abstract

Section: Methodsmentioning

confidence: 99%

Determination of three‐dimensional positions of known sparse objects from a single projection

Hoffmann

Esthappan

1997

Medical Physics

View full text Add to dashboard Cite

show abstract

“…Further, because the assumptions for a standard permutation test (von Storch and Zwiers 1999) are the same as those required for the Mann-Whitney U -test (as detailed in section 5), the test offers no advantage if the interest is only in testing the statistical signi cance of the area under the entire ROC curve. Of course, permutation tests can provide additional information about individual segments of the curve (although there are alternative methods of analysing speci c points on the curve (Metz 1978;Mann et al 1992)), and more sophisticated randomization tests can be designed if the assumptions for a standard test are invalid.…”

Section: The Roc Area and Mann-whitney U -Statistic (A) Continuous Fomentioning

confidence: 99%

“…If the sets of forecasts can be considered independent, a two-sample Kolmogorov-Smirnov test can be used to compare two ROCs (Gail and Green 1976;Campbell 1994), and makes no parametric assumptions about the distributions of the FPs (Sheskin 2000). Parametric alternatives have been suggested, based primarily on the t-test and Â 2 test (Swets et al 1961;Dorfman and Alf 1969;Metz 1978;Metz and Kronman 1980;Hanley and McNeil 1982), and have been applied in a weather forecasting context (Winston 1988). These techniques could similarly be applied for the comparison of ROL areas.…”

Section: Comparing Roc or Rol Areasmentioning

confidence: 99%

“…The technique has been used extensively in the elds of psychological and medical test evaluation (e.g. Green and Swets 1966;Swets 1973Swets , 1979Swets , 1988Swets , 1995Egan 1975;Metz 1978;Falmagne 1985;Begg 1991;Swets et al 2000) †, and is being applied increasingly in the atmospheric sciences (e.g. Mason 1982;Winston 1988;Buizza and Palmer 1998;Buizza et al 1998Buizza et al , 1999Graham et al 2000;Palmer et al 2000;Richardson 2000;Zhang and Casey 2000;Buizza 2001;Frogner and Iversen 2001;Mullen and Buizza 2001;Thorncroft and Pytharoulis 2001;Wandishin et al 2001;Wilks 2001;Zhu et al 2002).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Areas beneath the relative operating characteristics (ROC) and relative operating levels (ROL) curves: Statistical significance and interpretation

Mason

Graham

2002

Quart J Royal Meteoro Soc

733

527

View full text Add to dashboard Cite

SUMMARYThe areas beneath the relative (or receiver) operating characteristics (ROC) and relative operating levels (ROL) curves can be used as summary measures of forecast quality, but statistical signi cance tests for these areas are conducted infrequently in the atmospheric sciences. A development of signal-detection theory, the ROC curve has been widely applied in the medical and psychology elds where signi cance tests and relationships to other common statistical methods have been established and described. This valuable literature appears to be largely unknown to the atmospheric sciences where applications of ROC and related techniques are becoming more common. This paper presents a survey of that literature with a focus on the interpretation of the ROC area in the eld of forecast veri cation. We extend these foundations to demonstrate that similar principles can be applied to the interpretation and signi cance testing of the ROL area. It is shown that the ROC area is equivalent to the Mann-Whitney U -statistic testing the signi cance of forecast event probabilities for cases where events actually occurred with those where events did not occur. A similar derivation shows that the ROL area is equivalent to the Mann-Whitney U -statistic testing the magnitude of events with respect to whether or not an event has been forecast. Because the Mann-Whitney U -statistic follows a known probability distribution, under certain assumptions it can be used to de ne the statistical signi cance of ROC and ROL areas and for comparing the areas of competing forecasts. For large samples the signi cance of either measure can be accurately assessed using a normal-distribution approximation.

show abstract

“…Para tal feito, é criado, através dessa transformação, um conjunto unirrótulo multiclasses, na qual os subconjuntos, ou conjuntos únicos de rótulos, estão presentes no conjunto de treinamento e representados por valores unitários do atributo classe, a partir dos quais um classificador multiclasse é construído [35]. Neste método, cada subconjunto diferente de rótulos de L é considerado como uma única classe da nova tarefa de classificação unirrótulo.…”

Section: Label Powersetunclassified

Estratégias de Construções de Comitês de Classificadores Multirrótulos no Aprendizado Semissupervisionado Multidescrição

Silva

Santos

2017

RITA

View full text Add to dashboard Cite

Resumo: Classificação multirrótulo é um problema de aprendizado supervisionado no qual um objeto pode estar associado a múltiplas classes. Dentre os diferentes métodos de classificação multirrótulo destacam-se os métodos BR (Binary Relevance), LP (Label Powerset) e RAkEL (RAndom k-labELsets). O trabalho realizou um estudo sobre as construções de comitês de classificadores multirrótulos construídos através da aplicação de técnicas de aprendizado semissupervisionado multidescrição. Os comitês de classificadores utilizados nos experimentos foram o Bagging, Boosting e Stacking; como métodos de transformação do problema utilizamos os métodos BR, LP e Rakel; na classificação multirrótulo semissupervisionada multidescrição foi utilizado o Co-Training; foram aplicados cinco diferentes algoritmos como classificadores base: k-NN (k Vizinhos Mais Próximos), J48 (Algoritmo de Indução de Árvores de Decisão), SVM (Máquinas de Vetores Suporte), NB (Naive Bayes) e o JRip (Extended Repeated Incremental Pruning). Todos os experimentos utilizaram a metodologia de validação cruzada com 10 grupos (10-fold Cross-Validation) e o framework MULAN, o qual é implementado utilizando o WEKA. Para os tamanhos dos comitês de classificadores adotamos os valores 3, 5, 7 e 9. Para a análise dos resultados foi utilizado o teste estatístico de Wilcoxon. Ao final das análises experimentais, verificou-se que a abordagem semissupervisionado apresentou resultados competitivos em relação ao aprendizado supervisionado, uma vez que as duas abordagens utilizadas apresentaram resultados estatisticamente semelhantes.

show abstract

Abordagens para aprendizado semissupervisionado multirrótulo e hierárquico

Cited by 5 publications

References 70 publications

Determination of three‐dimensional positions of known sparse objects from a single projection

Determination of three‐dimensional positions of known sparse objects from a single projection

Areas beneath the relative operating characteristics (ROC) and relative operating levels (ROL) curves: Statistical significance and interpretation

Estratégias de Construções de Comitês de Classificadores Multirrótulos no Aprendizado Semissupervisionado Multidescrição

Contact Info

Product

Resources

About