Header logo is ei


1997


no image
Comparing support vector machines with Gaussian kernels to radial basis function classifiers

Schölkopf, B., Sung, K., Burges, C., Girosi, F., Niyogi, P., Poggio, T., Vapnik, V.

IEEE Transactions on Signal Processing, 45(11):2758-2765, November 1997 (article)

Abstract
The support vector (SV) machine is a novel type of learning machine, based on statistical learning theory, which contains polynomial classifiers, neural networks, and radial basis function (RBF) networks as special cases. In the RBF case, the SV algorithm automatically determines centers, weights, and threshold that minimize an upper bound on the expected test error. The present study is devoted to an experimental comparison of these machines with a classical approach, where the centers are determined by X-means clustering, and the weights are computed using error backpropagation. We consider three machines, namely, a classical RBF machine, an SV machine with Gaussian kernel, and a hybrid system with the centers determined by the SV method and the weights trained by error backpropagation. Our results show that on the United States postal service database of handwritten digits, the SV machine achieves the highest recognition accuracy, followed by the hybrid system. The SV approach is thus not only theoretically well-founded but also superior in a practical application.

Web DOI [BibTex]

1997

Web DOI [BibTex]


no image
The view-graph approach to visual navigation and spatial memory

Mallot, H., Franz, M., Schölkopf, B., Bülthoff, H.

In Artificial Neural Networks: ICANN ’97, pages: 751-756, (Editors: W Gerstner and A Germond and M Hasler and J-D Nicoud), Springer, Berlin, Germany, 7th International Conference on Artificial Neural Networks, October 1997 (inproceedings)

Abstract
This paper describes a purely visual navigation scheme based on two elementary mechanisms (piloting and guidance) and a graph structure combining individual navigation steps controlled by these mechanisms. In robot experiments in real environments, both mechanisms have been tested, piloting in an open environment and guidance in a maze with restricted movement opportunities. The results indicate that navigation and path planning can be brought about with these simple mechanisms. We argue that the graph of local views (snapshots) is a general and biologically plausible means of representing space and integrating the various mechanisms of map behaviour.

PDF PDF DOI [BibTex]

PDF PDF DOI [BibTex]


no image
Predicting time series with support vector machines

Müller, K., Smola, A., Rätsch, G., Schölkopf, B., Kohlmorgen, J., Vapnik, V.

In Artificial Neural Networks: ICANN’97, pages: 999-1004, (Editors: Schölkopf, B. , C.J.C. Burges, A.J. Smola), Springer, Berlin, Germany, 7th International Conference on Artificial Neural Networks , October 1997 (inproceedings)

Abstract
Support Vector Machines are used for time series prediction and compared to radial basis function networks. We make use of two different cost functions for Support Vectors: training with (i) an e insensitive loss and (ii) Huber's robust loss function and discuss how to choose the regularization parameters in these models. Two applications are considered: data from (a) a noisy (normal and uniform noise) Mackey Glass equation and (b) the Santa Fe competition (set D). In both cases Support Vector Machines show an excellent performance. In case (b) the Support Vector approach improves the best known result on the benchmark by a factor of 29%.

PDF DOI [BibTex]

PDF DOI [BibTex]


no image
Predicting time series with support vectur machines

Müller, K., Smola, A., Rätsch, G., Schölkopf, B., Kohlmorgen, J., Vapnik, V.

In Artificial neural networks: ICANN ’97, pages: 999-1004, (Editors: W Gerstner and A Germond and M Hasler and J-D Nicoud), Springer, Berlin, Germany, 7th International Conference on Artificial Neural Networks , October 1997 (inproceedings)

Abstract
Support Vector Machines are used for time series prediction and compared to radial basis function networks. We make use of two different cost functions for Support Vectors: training with (i) an e insensitive loss and (ii) Huber's robust loss function and discuss how to choose the regularization parameters in these models. Two applications are considered: data from (a) a noisy (normal and uniform noise) Mackey Glass equation and (b) the Santa Fe competition (set D). In both cases Support Vector Machines show an excellent performance. In case (b) the Support Vector approach improves the best known result on the benchmark by a factor of 29%.

PDF PDF DOI [BibTex]

PDF PDF DOI [BibTex]


no image
Kernel principal component analysis

Schölkopf, B., Smola, A., Müller, K.

In Artificial neural networks: ICANN ’97, LNCS, vol. 1327, pages: 583-588, (Editors: W Gerstner and A Germond and M Hasler and J-D Nicoud), Springer, Berlin, Germany, 7th International Conference on Artificial Neural Networks, October 1997 (inproceedings)

Abstract
A new method for performing a nonlinear form of Principal Component Analysis is proposed. By the use of integral operator kernel functions, one can efficiently compute principal components in highdimensional feature spaces, related to input space by some nonlinear map; for instance the space of all possible d-pixel products in images. We give the derivation of the method and present experimental results on polynomial feature extraction for pattern recognition.

PDF DOI [BibTex]

PDF DOI [BibTex]


no image
Homing by parameterized scene matching

Franz, M., Schölkopf, B., Bülthoff, H.

In Proceedings of the 4th European Conference on Artificial Life, (Eds.) P. Husbands, I. Harvey. MIT Press, Cambridge 1997, pages: 236-245, (Editors: P Husbands and I Harvey), MIT Press, Cambridge, MA, USA, 4th European Conference on Artificial Life (ECAL97), July 1997 (inproceedings)

Abstract
In visual homing tasks, animals as well as robots can compute their movements from the current view and a snapshot taken at a home position. Solving this problem exactly would require knowledge about the distances to visible landmarks, information, which is not directly available to passive vision systems. We propose a homing scheme that dispenses with accurate distance information by using parameterized disparity fields. These are obtained from an approximation that incorporates prior knowledge about perspective distortions of the visual environment. A mathematical analysis proves that the approximation does not prevent the scheme from approaching the goal with arbitrary accuracy. Mobile robot experiments are used to demonstrate the practical feasibility of the approach.

PDF [BibTex]

PDF [BibTex]


no image
Das Spiel mit dem künstlichen Leben.

Schölkopf, B.

Frankfurter Allgemeine Zeitung, Wissenschaftsbeilage, June 1997 (misc)

[BibTex]

[BibTex]


no image
Improving the accuracy and speed of support vector learning machines

Burges, C., Schölkopf, B.

In Advances in Neural Information Processing Systems 9, pages: 375-381, (Editors: M Mozer and MJ Jordan and T Petsche), MIT Press, Cambridge, MA, USA, Tenth Annual Conference on Neural Information Processing Systems (NIPS), May 1997 (inproceedings)

Abstract
Support Vector Learning Machines (SVM) are finding application in pattern recognition, regression estimation, and operator inversion for illposed problems . Against this very general backdrop any methods for improving the generalization performance, or for improving the speed in test phase of SVMs are of increasing interest. In this paper we combine two such techniques on a pattern recognition problem The method for improving generalization performance the "virtual support vector" method does so by incorporating known invariances of the problem This method achieves a drop in the error rate on 10.000 NIST test digit images of 1,4 % to 1 %. The method for improving the speed (the "reduced set" method) does so by approximating the support vector decision surface. We apply this method to achieve a factor of fifty speedup in test phase over the virtual support vector machine The combined approach yields a machine which is both 22 times faster than the original machine, and which has better generalization performance achieving 1,1 % error . The virtual support vector method is applicable to any SVM problem with known invariances The reduced set method is applicable to any support vector machine .

PDF Web [BibTex]

PDF Web [BibTex]


no image
Homing by parameterized scene matching

Franz, M., Schölkopf, B., Bülthoff, H.

(46), Max Planck Institute for Biological Cybernetics, Tübingen, Germany, Febuary 1997 (techreport)

Abstract
In visual homing tasks, animals as well as robots can compute their movements from the current view and a snapshot taken at a home position. Solving this problem exactly would require knowledge about the distances to visible landmarks, information, which is not directly available to passive vision systems. We propose a homing scheme that dispenses with accurate distance information by using parameterized disparity fields. These are obtained from an approximation that incorporates prior knowledge about perspective distortions of the visual environment. A mathematical analysis proves that the approximation does not prevent the scheme from approaching the goal with arbitrary accuracy. Mobile robot experiments are used to demonstrate the practical feasibility of the approach.

[BibTex]

[BibTex]


no image
Learning view graphs for robot navigation

Franz, M., Schölkopf, B., Georg, P., Mallot, H., Bülthoff, H.

In Proceedings of the 1st Intl. Conf. on Autonomous Agents, pages: 138-147, (Editors: Johnson, W.L.), ACM Press, New York, NY, USA, First International Conference on Autonomous Agents (AGENTS '97), Febuary 1997 (inproceedings)

Abstract
We present a purely vision-based scheme for learning a parsimonious representation of an open environment. Using simple exploration behaviours, our system constructs a graph of appropriately chosen views. To navigate between views connected in the graph, we employ a homing strategy inspired by findings of insect ethology. Simulations and robot experiments demonstrate the feasibility of the proposed approach.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Masking by plaid patterns is not explained by adaptation, simple contrast gain-control or distortion products

Wichmann, F., Tollin, D.

Investigative Ophthamology and Visual Science, 38 (4), pages: S631, 1997 (poster)

[BibTex]

[BibTex]


no image
Masking by plaid patterns: spatial frequency tuning and contrast dependency

Wichmann, F., Tollin, D.

OSA Conference Program, pages: 97, 1997 (poster)

Abstract
The detectability of horizontally orientated sinusoidal signals at different spatial-frequencies was measured in standard 2AFC - tasks in the presence of two-component plaid patterns of different orientation and contrast. The shape of the resulting masking surface provides insight into, and constrains models of, the underlying masking mechanisms.

[BibTex]

[BibTex]


no image
ATM-dependent telomere loss in aging human diploid fibroblasts and DNA damage lead to the post-translational activation of p53 protein involving poly(ADP-ribose) polymerase.

Vaziri, H., MD, .., RC, .., Davison, T., YS, .., CH, .., GG, .., Benchimol, S.

The European Molecular Biology Organization Journal, 16(19):6018-6033, 1997 (article)

Web [BibTex]

Web [BibTex]


no image
Support vector learning

Schölkopf, B.

pages: 173, Oldenbourg, München, Germany, 1997, Zugl.: Berlin, Techn. Univ., Diss., 1997 (book)

PDF GZIP [BibTex]

PDF GZIP [BibTex]

1995


no image
View-based cognitive map learning by an autonomous robot

Mallot, H., Bülthoff, H., Georg, P., Schölkopf, B., Yasuhara, K.

In Proceedings International Conference on Artificial Neural Networks, vol. 2, pages: 381-386, (Editors: Fogelman-Soulié, F.), EC2, Paris, France, Conférence Internationale sur les Réseaux de Neurones Artificiels (ICANN '95), October 1995 (inproceedings)

Abstract
This paper presents a view-based approach to map learning and navigation in mazes. By means of graph theory we have shown that the view-graph is a sufficient representation for map behaviour such as path planning. A neural network for unsupervised learning of the view-graph from sequences of views is constructed. We use a modified Kohonen (1988) learning rule that transforms temporal sequence (rather than featural similarity) into connectedness. In the main part of the paper, we present a robot implementation of the scheme. The results show that the proposed network is able to support map behaviour in simple environments.

PDF [BibTex]

1995

PDF [BibTex]


no image
Extracting support data for a given task

Schölkopf, B., Burges, C., Vapnik, V.

In First International Conference on Knowledge Discovery & Data Mining (KDD-95), pages: 252-257, (Editors: UM Fayyad and R Uthurusamy), AAAI Press, Menlo Park, CA, USA, August 1995 (inproceedings)

Abstract
We report a novel possibility for extracting a small subset of a data base which contains all the information necessary to solve a given classification task: using the Support Vector Algorithm to train three different types of handwritten digit classifiers, we observed that these types of classifiers construct their decision surface from strongly overlapping small (k: 4%) subsets of the data base. This finding opens up the possibiiity of compressing data bases significantly by disposing of the data which is not important for the solution of a given task. In addition, we show that the theory allows us to predict the classifier that will have the best generalization ability, based solely on performance on the training set and characteristics of the learning machines. This finding is important for cases where the amount of available data is limited.

PDF [BibTex]

PDF [BibTex]


no image
View-Based Cognitive Mapping and Path Planning

Schölkopf, B., Mallot, H.

Adaptive Behavior, 3(3):311-348, January 1995 (article)

Abstract
This article presents a scheme for learning a cognitive map of a maze from a sequence of views and movement decisions. The scheme is based on an intermediate representation called the view graph, whose nodes correspond to the views whereas the labeled edges represent the movements leading from one view to another. By means of a graph theoretical reconstruction method, the view graph is shown to carry complete information on the topological and directional structure of the maze. Path planning can be carried out directly in the view graph without actually performing this reconstruction. A neural network is presented that learns the view graph during a random exploration of the maze. It is based on an unsupervised competitive learning rule translating temporal sequence (rather than similarity) of views into connectedness in the network. The network uses its knowledge of the topological and directional structure of the maze to generate expectations about which views are likely to be encountered next, improving the view-recognition performance. Numerical simulations illustrate the network's ability for path planning and the recognition of views degraded by random noise. The results are compared to findings of behavioral neuroscience.

Web DOI [BibTex]

Web DOI [BibTex]


no image
Suppression and creation of chaos in a periodically forced Lorenz system.

Franz, MO., Zhang, MH.

Physical Review, E 52, pages: 3558-3565, 1995 (article)

Abstract
Periodic forcing is introduced into the Lorenz model to study the effects of time-dependent forcing on the behavior of the system. Such a nonautonomous system stays dissipative and has a bounded attracting set which all trajectories finally enter. The possible kinds of attracting sets are restricted to periodic orbits and strange attractors. A large-scale survey of parameter space shows that periodic forcing has mainly three effects in the Lorenz system depending on the forcing frequency: (i) Fixed points are replaced by oscillations around them; (ii) resonant periodic orbits are created both in the stable and the chaotic region; (iii) chaos is created in the stable region near the resonance frequency and in periodic windows. A comparison to other studies shows that part of this behavior has been observed in simulations of higher truncations and real world experiments. Since very small modulations can already have a considerable effect, this suggests that periodic processes such as annual or diurnal cycles should not be omitted even in simple climate models.

[BibTex]

[BibTex]


no image
A New Method for Constructing Artificial Neural Networks

Vapnik, V., Burges, C., Schölkopf, B.

AT & T Bell Laboratories, 1995 (techreport)

[BibTex]

[BibTex]


no image
Image segmentation from motion: just the loss of high-spatial-frequency content ?

Wichmann, F., Henning, G.

Perception, 24, pages: S19, 1995 (poster)

Abstract
The human contrast sensitivity function (CSF) is bandpass for stimuli of low temporal frequency but, for moving stimuli, results in a low-pass CSF with large high spatial-frequency losses. Thus the high spatial-frequency content of images moving on the retina cannot be seen; motion perception could be facilitated by, or even be based on, the selective loss of high spatial-frequency content. 2-AFC image segmentation experiments were conducted with segmentation based on motion or on form. In the latter condition, the form difference mirrored that produced by moving stimuli. This was accomplished by generating stimulus elements which were spectrally either broadband or low-pass. For the motion used, the spectral difference between static broadband and static low-pass elements matched the spectral difference between moving and static broadband elements. On the hypothesis that segmentation from motion is based on the detection of regions devoid of high spatial-frequencies, both tasks should be similarly difficult for human observers. However, neither image segmentation (nor, incidentally, motion detection) was sensitive to the high spatial-frequency content of the stimuli. Thus changes in perceptual form produced by moving stimuli appear not to be used as a cue for image segmentation.

[BibTex]