Results 1  10
of
34
Multicategory Classification by Support Vector Machines
 Computational Optimizations and Applications
, 1999
"... We examine the problem of how to discriminate between objects of three or more classes. Specifically, we investigate how twoclass discrimination methods can be extended to the multiclass case. We show how the linear programming (LP) approaches based on the work of Mangasarian and quadratic programm ..."
Abstract

Cited by 56 (0 self)
 Add to MetaCart
We examine the problem of how to discriminate between objects of three or more classes. Specifically, we investigate how twoclass discrimination methods can be extended to the multiclass case. We show how the linear programming (LP) approaches based on the work of Mangasarian and quadratic programming (QP) approaches based on Vapnik's Support Vector Machines (SVM) can be combined to yield two new approaches to the multiclass problem. In LP multiclass discrimination, a single linear program is used to construct a piecewise linear classification function. In our proposed multiclass SVM method, a single quadratic program is used to construct a piecewise nonlinear classification function. Each piece of this function can take the form of a polynomial, radial basis function, or even a neural network. For the k > 2 class problems, the SVM method as originally proposed required the construction of a twoclass SVM to separate each class from the remaining classes. Similarily, k twoclass linear programs can be used for the multiclass problem. We performed an empirical study of the original LP method, the proposed k LP method, the proposed single QP method and the original k QP methods. We discuss the advantages and disadvantages of each approach. 1 1
Where are linear feature extraction methods applicable
 IEEE Trans. Pattern Anal. Mach. Intell
, 2005
"... Abstract—A fundamental problem in computer vision and pattern recognition is to determine where and, most importantly, why a given technique is applicable. This is not only necessary because it helps us decide which techniques to apply at each given time. Knowing why current algorithms cannot be app ..."
Abstract

Cited by 34 (14 self)
 Add to MetaCart
Abstract—A fundamental problem in computer vision and pattern recognition is to determine where and, most importantly, why a given technique is applicable. This is not only necessary because it helps us decide which techniques to apply at each given time. Knowing why current algorithms cannot be applied facilitates the design of new algorithms robust to such problems. In this paper, we report on a theoretical study that demonstrates where and why generalized eigenbased linear equations do not work. In particular, we show that when the smallest angle between the ith eigenvector given by the metric to be maximized and the first i eigenvectors given by the metric to be minimized is close to zero, our results are not guaranteed to be correct. Several properties of such models are also presented. For illustration, we concentrate on the classical applications of classification and feature extraction. We also show how we can use our findings to design more robust algorithms. We conclude with a discussion on the broader impacts of our results. Index Terms—Feature extraction, generalized eigenvalue decomposition, performance evaluation, classifiers, pattern recognition. æ 1
Least absolute shrinkage is equivalent to quadratic penalization
 of Perspectives in Neural Computing
, 1998
"... Adaptive ridge is a special form of ridge regression, balancing the quadratic penalization on each parameter of the model. This paper shows the equivalence between adaptive ridge and lasso (least absolute shrinkage and selection operator). This equivalence states that both procedures produce the sam ..."
Abstract

Cited by 29 (7 self)
 Add to MetaCart
Adaptive ridge is a special form of ridge regression, balancing the quadratic penalization on each parameter of the model. This paper shows the equivalence between adaptive ridge and lasso (least absolute shrinkage and selection operator). This equivalence states that both procedures produce the same estimate. Least absolute shrinkage can thus be viewed as a particular quadratic penalization. From this observation, we derive an EM algorithm to compute the lasso solution. We finally present a series of applications of this type of algorithm in regression problems: kernel regression, additive modeling and neural net training. 1
Nearest Neighbors in Random Subspaces
 Lecture Notes in Computer Science: Advances in Pattern Recognition
, 1998
"... Recent studies have shown that the random subspace method can be used to create multiple independent treeclassifiers that can be combined to improve accuracy. We apply the procedure to knearestneighbor classifiers and show that it can achieve similar results. We examine the effects of several ..."
Abstract

Cited by 24 (1 self)
 Add to MetaCart
Recent studies have shown that the random subspace method can be used to create multiple independent treeclassifiers that can be combined to improve accuracy. We apply the procedure to knearestneighbor classifiers and show that it can achieve similar results. We examine the effects of several parameters of the method by experiments using data from a digit recognition problem. We show that the combined accuracies follow a trend of increase with increasing number of component classifiers, and that with an appropriate subspace dimensionality, the method can be superior to simple knearestneighbor classification.
Outcomes of the equivalence of adaptive ridge with least absolute shrinkage
 Eds.), Advances in Neural Information Processing Systems
, 1998
"... Adaptive Ridge is a special form of Ridge regression, balancing the quadratic penalization on each parameter of the model. It was shown to be equivalent to Lasso (least absolute shrinkage and selection operator), in the sense that both procedures produce the same estimate. Lasso can thus be viewed a ..."
Abstract

Cited by 14 (4 self)
 Add to MetaCart
Adaptive Ridge is a special form of Ridge regression, balancing the quadratic penalization on each parameter of the model. It was shown to be equivalent to Lasso (least absolute shrinkage and selection operator), in the sense that both procedures produce the same estimate. Lasso can thus be viewed as a particular quadratic penalizer. From this observation, we derive a fixed point algorithm to compute the Lasso solution. The analogy provides also a new hyperparameter for tuning effectively the model complexity. We finally present a series of possible extensions of lasso performing sparse regression in kernel smoothing, additive modeling and neural net training. 1
Directed search in a 3D objects database using SVM", HewlettPackard Research Report HPL200020R1
, 2000
"... search in databases, moments, support vector machine, quadratic programming This paper introduces a contentbased search algorithm for a database of 3D objects. The search is performed by giving an example object, and looking for similar ones in the database. The search system result is given as sev ..."
Abstract

Cited by 11 (0 self)
 Add to MetaCart
search in databases, moments, support vector machine, quadratic programming This paper introduces a contentbased search algorithm for a database of 3D objects. The search is performed by giving an example object, and looking for similar ones in the database. The search system result is given as several nearest neighbor objects. The weighted Euclidean distance between a sequence of normalized moments is used to measure the similarity between objects. The moments are estimated based on uniformly distributed random generation of 3D points on the objects ' surface. An important feature of the search system is the proposed iterative refinement algorithm. Marking successful and failure decisions on previous results, this user's feedback is used to adapt the weights of the distance measure. The adaptation causes the successful objects to become nearer, and the failure decisions to become distant. Training the distance measure is done using the wellknown SVM algorithm, which introduces robustness to the weight parameters. The above process may be repeated several times, accumulating 'Good ' and 'Bad ' examples and updating the distance measure to discriminate between the two with a maximal margin. This way, for each search and for each different user applying it, a different measure of similarity that reflects the user's desires and the searched object properties, is obtained. Simulations done on a database of more than 500 objects show promising results. It is shown that with only 23 such refinement iterations, the search results are successfully directed towards better suited 3D objects.
Virtual screens for ligands of orphan G proteincoupled receptors
 J. Chem. Inf. Model. 2005
"... Supporting Information Links to the 4 articles that cite this article, as of the time of this article download Access to high resolution figures Links to articles and content related to this article Copyright permission to reproduce figures and/or text from this article ..."
Abstract

Cited by 11 (0 self)
 Add to MetaCart
Supporting Information Links to the 4 articles that cite this article, as of the time of this article download Access to high resolution figures Links to articles and content related to this article Copyright permission to reproduce figures and/or text from this article
Tracking Body and Hands For Gesture Recognition: NATOPS Aircraft Handling Signals Database
 In FG
, 2011
"... Abstract — We present a unified framework for body and hand tracking, the output of which can be used for understanding simultaneously performed bodyandhand gestures. The framework uses a stereo camera to collect 3D images, and tracks body and hand together, combining various existing techniques t ..."
Abstract

Cited by 4 (3 self)
 Add to MetaCart
Abstract — We present a unified framework for body and hand tracking, the output of which can be used for understanding simultaneously performed bodyandhand gestures. The framework uses a stereo camera to collect 3D images, and tracks body and hand together, combining various existing techniques to make tracking tasks efficient. In addition, we introduce a multisignal gesture database: the NATOPS aircraft handling signals. Unlike previous gesture databases, this data requires knowledge about both body and hand in order to distinguish gestures. It is also focused on a clearly defined gesture vocabulary from a realworld scenario that has been refined over many years. The database includes 24 bodyandhand gestures, and provides both gesture video clips and the body and hand features we extracted. I.
Segmenting pointsampled surfaces
, 2010
"... Extracting features from pointbased representations of geometric surface models is becoming increasingly important for purposes such as model classification, matching, and exploration. In an earlier paper, we proposed a multiphase segmentation process to identify elongated features in pointsample ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
Extracting features from pointbased representations of geometric surface models is becoming increasingly important for purposes such as model classification, matching, and exploration. In an earlier paper, we proposed a multiphase segmentation process to identify elongated features in pointsampled surface models without the explicit construction of a mesh or other surface representation. The preliminary results demonstrated the strength and potential of the segmentation process, but the resulting segmentations were still of low quality, and the segmentation process could be slow. In this paper, we describe several algorithmic improvements to overcome the shortcomings of the segmentation process. To demonstrate the improved quality of the segmentation and the superior time efficiency of the new segmentation process, we present segmentation results obtained for various pointsampled surface models. We also discuss an application of our segmentation process to extract
BlackBox Software Sensor Design for Environmental Monitoring
 In: ICANN'98, SpringerVerlag
, 1998
"... Software sensor design consists in building a model to estimate an unknownquantity, with error bars, using other available measurements. In the environmental domain, due to a lack of physical model, nonlinearities, and unknown time dependencies, blackbox modelling is required. An application in ri ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
Software sensor design consists in building a model to estimate an unknownquantity, with error bars, using other available measurements. In the environmental domain, due to a lack of physical model, nonlinearities, and unknown time dependencies, blackbox modelling is required. An application in river water quality monitoring illustrates a neural network based methodology. All stages of the method are described from data cleaning, and model selection, predictor estimation and prediction validity assessment. The originality of the approach is that it provides automatically an estimation of inputs relevance in merging the input selection and prediction estimation steps. 1 Introduction Environmental monitoringrequires valid measurements but sensors are often either expensive or unreliable. The development of software sensors is a major issue for the next generation of monitoring devices. The problem is to build a model capable of giving an estimate of the quantity of interest, with a co...