Publications | Max Planck Institute for Intelligent Systems

419 results (View BibTeX file of all listed publications)

2006

Task-based and stable telenanomanipulation in a nanoscale virtual environment

IEEE Transactions on automation science and engineering, 3(3):240-247, IEEE, 2006 (article)

2006

pi Kim, S., Sitti, M. Task-based and stable telenanomanipulation in a nanoscale virtual environment IEEE Transactions on automation science and engineering, 3(3):240-247, IEEE, 2006 (article)

[BibTex]

Drawing suspended polymer micro-/nanofibers using glass micropipettes

Nain, A. S., Wong, J. C., Amon, C., Sitti, M.

Applied Physics Letters, 89(18):183105, AIP, 2006 (article)

[BibTex]

pi Nain, A. S., Wong, J. C., Amon, C., Sitti, M. Drawing suspended polymer micro-/nanofibers using glass micropipettes Applied Physics Letters, 89(18):183105, AIP, 2006 (article)

[BibTex]

Dynamic Hebbian learning in adaptive frequency oscillators

Righetti, L., Buchli, J., Ijspeert, A.

Physica D: Nonlinear Phenomena, 216(2):269-281, 2006 (article)

Abstract

Nonlinear oscillators are widely used in biology, physics and engineering for modeling and control. They are interesting because of their synchronization properties when coupled to other dynamical systems. In this paper, we propose a learning rule for oscillators which adapts their frequency to the frequency of any periodic or pseudo-periodic input signal. Learning is done in a dynamic way: it is part of the dynamical system and not an offline process. An interesting property of our model is that it is easily generalizable to a large class of oscillators, from phase oscillators to relaxation oscillators and strange attractors with a generic learning rule. One major feature of our learning rule is that the oscillators constructed can adapt their frequency without any signal processing or the need to specify a time window or similar free parameters. All the processing is embedded in the dynamics of the adaptive oscillator. The convergence of the learning is proved for the Hopf oscillator, then numerical experiments are carried out to explore the learning capabilities of the system. Finally, we generalize the learning rule to non-harmonic oscillators like relaxation oscillators and strange attractors.

link (url) DOI [BibTex]

mg Righetti, L., Buchli, J., Ijspeert, A. Dynamic Hebbian learning in adaptive frequency oscillators Physica D: Nonlinear Phenomena, 216(2):269-281, 2006 (article)

link (url) DOI [BibTex]

Theoretical justification of ground-state moment analysis of magnetic dichroic x-ray absorption spectra for 3d transition metals

Dörfler, F., Fähnle, M.

{Physical Review B}, 74, 2006 (article)

Abstract

{(6 pages)}

mms

[BibTex]

mms Dörfler, F., Fähnle, M. Theoretical justification of ground-state moment analysis of magnetic dichroic x-ray absorption spectra for 3d transition metals {Physical Review B}, 74, 2006 (article)

[BibTex]

Vortex dynamics in coupled ferromagnetic multilayer structures

Chou, K., Puzic, A., Stoll, H., Schütz, G., Van Waeyenberge, B., Tyliszczak, T., Rott, K., Reiss, G., Brückl, H., Neudecker, I., Weiss, D., Back, C. H.

{Journal of Applied Physics}, 99, 2006 (article)

mms

[BibTex]

mms Chou, K., Puzic, A., Stoll, H., Schütz, G., Van Waeyenberge, B., Tyliszczak, T., Rott, K., Reiss, G., Brückl, H., Neudecker, I., Weiss, D., Back, C. H. Vortex dynamics in coupled ferromagnetic multilayer structures {Journal of Applied Physics}, 99, 2006 (article)

[BibTex]

Nonlinear magnetoelastic behavior of the bcc phases of Co an Ni: importance of third-order contributions for bcc Ni

Komelj, M., Fähnle, M.

{Physical Review B}, 73, 2006 (article)

mms

[BibTex]

mms Komelj, M., Fähnle, M. Nonlinear magnetoelastic behavior of the bcc phases of Co an Ni: importance of third-order contributions for bcc Ni {Physical Review B}, 73, 2006 (article)

[BibTex]

Electrical transport, magnetic and thermal properties of icosahedral Al-Pd-Mn quasicrystals

Poddar, A., Das, S., Plachke, D., Carstanjen, H.D.

{Journal of Magnetism and Magnetic Materials}, 300(2):263-272, 2006 (article)

mms

[BibTex]

mms Poddar, A., Das, S., Plachke, D., Carstanjen, H.D. Electrical transport, magnetic and thermal properties of icosahedral Al-Pd-Mn quasicrystals {Journal of Magnetism and Magnetic Materials}, 300(2):263-272, 2006 (article)

[BibTex]

Aqueous solution deposition of indium hydroxide and indium oxide columnar type thin films

Qiu, Y., Gerstel, P., Jiang, L., Lipowsky, P., Pitta Bauermann, L., Bill, J.

{International Journal of Materials Research}, 97(6):808-811, 2006 (article)

mms

[BibTex]

mms Qiu, Y., Gerstel, P., Jiang, L., Lipowsky, P., Pitta Bauermann, L., Bill, J. Aqueous solution deposition of indium hydroxide and indium oxide columnar type thin films {International Journal of Materials Research}, 97(6):808-811, 2006 (article)

[BibTex]

Enhanced orbital magnetism in Fe50Pt50 nanoparticles

Antoniak, C., Lindner, J., Spasova, M., Sudfeld, D., Acet, M., Farle, M., Fauth, K., Wiedwald, U., Boyen, H.-G., Ziemann, P., Wilhelm, F., Rogalev, A., Sun, S.

{Physical Review Letters}, 97, 2006 (article)

mms

[BibTex]

mms Antoniak, C., Lindner, J., Spasova, M., Sudfeld, D., Acet, M., Farle, M., Fauth, K., Wiedwald, U., Boyen, H.-G., Ziemann, P., Wilhelm, F., Rogalev, A., Sun, S. Enhanced orbital magnetism in Fe50Pt50 nanoparticles {Physical Review Letters}, 97, 2006 (article)

[BibTex]

Approximate nearest neighbor regression in very high dimensions

Vijayakumar, S., DSouza, A., Schaal, S.

In Nearest-Neighbor Methods in Learning and Vision, pages: 103-142, (Editors: Shakhnarovich, G.;Darrell, T.;Indyk, P.), Cambridge, MA: MIT Press, 2006, clmc (inbook)

link (url) [BibTex]

am Vijayakumar, S., DSouza, A., Schaal, S. Approximate nearest neighbor regression in very high dimensions In Nearest-Neighbor Methods in Learning and Vision, pages: 103-142, (Editors: Shakhnarovich, G.;Darrell, T.;Indyk, P.), Cambridge, MA: MIT Press, 2006, clmc (inbook)

link (url) [BibTex]

Biologically inspired polymer microfibers with spatulate tips as repeatable fibrillar adhesives

Kim, S., Sitti, M.

Applied Physics Letters, 89(26):261911-261911, AIP, 2006 (article)

Project Page [BibTex]

pi Kim, S., Sitti, M. Biologically inspired polymer microfibers with spatulate tips as repeatable fibrillar adhesives Applied Physics Letters, 89(26):261911-261911, AIP, 2006 (article)

Project Page [BibTex]

Converting polycrystals into single crystals - Selective grain growth by high-energy ion bombardment

Olliges, S., Gruber, P., Bardill, A., Ehrler, D., Carstanjen, H. D., Spolenak, R.

{Acta Materialia}, 54(20):5393-5399, 2006 (article)

mms

[BibTex]

mms Olliges, S., Gruber, P., Bardill, A., Ehrler, D., Carstanjen, H. D., Spolenak, R. Converting polycrystals into single crystals - Selective grain growth by high-energy ion bombardment {Acta Materialia}, 54(20):5393-5399, 2006 (article)

[BibTex]

Topological k-space refinement of the configurational energy of alloys

Shchyglo, O., Bugaev, V.N., Drautz, R., Udyanskyy, A., Reichert, H., Dosch, H.

{Physical Review B}, 72, 2006 (article)

mms

[BibTex]

mms Shchyglo, O., Bugaev, V.N., Drautz, R., Udyanskyy, A., Reichert, H., Dosch, H. Topological k-space refinement of the configurational energy of alloys {Physical Review B}, 72, 2006 (article)

[BibTex]

Synthesis of MgB2 films in Mg vapour flow and their characterization

Matveev, A. T., Albrecht, J., Konuma, M., Christiani, G., Krockenberger, Y., Starke, U., Schütz, G., Habermeier, H.

{Superconductor Science and Technology}, 19, pages: 299-305, 2006 (article)

mms

[BibTex]

mms Matveev, A. T., Albrecht, J., Konuma, M., Christiani, G., Krockenberger, Y., Starke, U., Schütz, G., Habermeier, H. Synthesis of MgB2 films in Mg vapour flow and their characterization {Superconductor Science and Technology}, 19, pages: 299-305, 2006 (article)

[BibTex]

Analysis of switching times of inhomogeneous magnetization processes in thin platelets

Goll, D., Schütz, G., Kronmüller, H.

{Physica B: Condensed Matter}, 372(1-2):273-276, 2006 (article)

mms

[BibTex]

mms Goll, D., Schütz, G., Kronmüller, H. Analysis of switching times of inhomogeneous magnetization processes in thin platelets {Physica B: Condensed Matter}, 372(1-2):273-276, 2006 (article)

[BibTex]

The initial oxidation of magnesium: an in situ study with XPS, HERDA and ellipsometry.

Kurth, M., Graat, P. C. J., Carstanjen, H. D., Mittemeijer, E. J.

{Surface and Interface Analysis}, 38, pages: 931-940, 2006 (article)

mms

[BibTex]

mms Kurth, M., Graat, P. C. J., Carstanjen, H. D., Mittemeijer, E. J. The initial oxidation of magnesium: an in situ study with XPS, HERDA and ellipsometry. {Surface and Interface Analysis}, 38, pages: 931-940, 2006 (article)

[BibTex]

Temperature influence of the faceting of \sum3 and \sum9 grain boundaries in Cu

Straumal, B. B., Polyakov, S. A., Mittemeijer, E. J.

{Acta Materialia}, 54, pages: 167-172, 2006 (article)

mms

[BibTex]

mms Straumal, B. B., Polyakov, S. A., Mittemeijer, E. J. Temperature influence of the faceting of \sum3 and \sum9 grain boundaries in Cu {Acta Materialia}, 54, pages: 167-172, 2006 (article)

[BibTex]

Coercivity of 2 : 17 based permanent magnets

Kronmüller, H., Goll, D.

{Journal of the Iron and Steel Research International}, 13(Suppl. 1):39-47, 2006 (article)

mms

[BibTex]

mms Kronmüller, H., Goll, D. Coercivity of 2 : 17 based permanent magnets {Journal of the Iron and Steel Research International}, 13(Suppl. 1):39-47, 2006 (article)

[BibTex]

Hydrogen adsorption in a nickel based coordination polymer with open metal sites in the cylindrical cavities of the desolvated framework

Dietzel, P. D. C., Panella, B., Hirscher, M., Blom, R., Fjellvag, H.

{Chemical Communications}, [2006], pages: 959-961, 2006 (article)

mms

DOI [BibTex]

mms Dietzel, P. D. C., Panella, B., Hirscher, M., Blom, R., Fjellvag, H. Hydrogen adsorption in a nickel based coordination polymer with open metal sites in the cylindrical cavities of the desolvated framework {Chemical Communications}, [2006], pages: 959-961, 2006 (article)

DOI [BibTex]

Magnetic ground states and the role of vortices in ferromagnetic hollow nanospheres

Goll, D., Macke, S., Berkowitz, A., Bertram, H. N.

{Physica B: Condensed Matter}, 372(1-2):282-285, 2006 (article)

mms

[BibTex]

mms Goll, D., Macke, S., Berkowitz, A., Bertram, H. N. Magnetic ground states and the role of vortices in ferromagnetic hollow nanospheres {Physica B: Condensed Matter}, 372(1-2):282-285, 2006 (article)

[BibTex]

Two-dimensional vision-based autonomous microparticle manipulation using a nanoprobe

Pawashe, C., Sitti, M.

Journal of Micromechatronics, 3(3):285-306, Brill, 2006 (article)

[BibTex]

pi Pawashe, C., Sitti, M. Two-dimensional vision-based autonomous microparticle manipulation using a nanoprobe Journal of Micromechatronics, 3(3):285-306, Brill, 2006 (article)

[BibTex]

Engineering Entrainment and Adaptation in Limit Cycle Systems – From biological inspiration to applications in robotics

Buchli, J., Righetti, L., Ijspeert, A.

Biological Cybernetics, 95(6):645-664, December 2006 (article)

Abstract

Periodic behavior is key to life and is observed in multiple instances and at multiple time scales in our metabolism, our natural environment, and our engineered environment. A natural way of modeling or generating periodic behavior is done by using oscillators, i.e., dynamical systems that exhibit limit cycle behavior. While there is extensive literature on methods to analyze such dynamical systems, much less work has been done on methods to synthesize an oscillator to exhibit some specific desired characteristics. The goal of this article is twofold: (1) to provide a framework for characterizing and designing oscillators and (2) to review how classes of well-known oscillators can be understood and related to this framework. The basis of the framework is to characterize oscillators in terms of their fundamental temporal and spatial behavior and in terms of properties that these two behaviors can be designed to exhibit. This focus on fundamental properties is important because it allows us to systematically compare a large variety of oscillators that might at first sight appear very different from each other. We identify several specifications that are useful for design, such as frequency-locking behavior, phase-locking behavior, and specific output signal shape. We also identify two classes of design methods by which these specifications can be met, namely offline methods and online methods. By relating these specifications to our framework and by presenting several examples of how oscillators have been designed in the literature, this article provides a useful methodology and toolbox for designing oscillators for a wide range of purposes. In particular, the focus on synthesis of limit cycle dynamical systems should be useful both for engineering and for computational modeling of physical or biological phenomena.

link (url) DOI [BibTex]

mg Buchli, J., Righetti, L., Ijspeert, A. Engineering Entrainment and Adaptation in Limit Cycle Systems – From biological inspiration to applications in robotics Biological Cybernetics, 95(6):645-664, December 2006 (article)

link (url) DOI [BibTex]

Pulsed-field-gradient NMR study of hydrogen diffusivity in random b.c.c. alloys VyTa1-y

Grinberg, F., Majer, G., Skripov, A. N.

{Journal of Alloys and Compounds}, 425(1-2):24-27, 2006 (article)

mms

[BibTex]

mms Grinberg, F., Majer, G., Skripov, A. N. Pulsed-field-gradient NMR study of hydrogen diffusivity in random b.c.c. alloys VyTa1-y {Journal of Alloys and Compounds}, 425(1-2):24-27, 2006 (article)

[BibTex]

Breathing Fermi surface model for noncollinear magnetization: a generalization of the Gilbert equation

Fähnle, M., Steiauf, D.

{Physical Review B}, 73, 2006 (article)

mms

[BibTex]

mms Fähnle, M., Steiauf, D. Breathing Fermi surface model for noncollinear magnetization: a generalization of the Gilbert equation {Physical Review B}, 73, 2006 (article)

[BibTex]

Reconstruction of brass for tongues and shallots from baroque organs

Baretzky, B., Friesel, M., Petelin, A., Shotanov, A., Straumal, B.

{Defect and Diffusion Forum}, 258-260, pages: 397-402, 2006 (article)

mms

DOI [BibTex]

mms Baretzky, B., Friesel, M., Petelin, A., Shotanov, A., Straumal, B. Reconstruction of brass for tongues and shallots from baroque organs {Defect and Diffusion Forum}, 258-260, pages: 397-402, 2006 (article)

DOI [BibTex]

Softening of nanostructured Al-Zn and Al-Mg alloys after severe plastic deformation

Mazilkin, A.A., Straumal, B.B., Rabkin, E., Baretzky, B., Enders, S., Protasova, S.G., Kogtenkova, O.A., Valiev, R.Z.

{Acta Materialia}, 54, pages: 3933-3939, 2006 (article)

mms

[BibTex]

mms Mazilkin, A.A., Straumal, B.B., Rabkin, E., Baretzky, B., Enders, S., Protasova, S.G., Kogtenkova, O.A., Valiev, R.Z. Softening of nanostructured Al-Zn and Al-Mg alloys after severe plastic deformation {Acta Materialia}, 54, pages: 3933-3939, 2006 (article)

[BibTex]

Comment on "Spin and orbital magnetic moments of Fe3O4

Goering, E., Lafkioti, M., Gold, S.

{Physical Review Letters}, 96(3), 2006 (article)

mms

[BibTex]

mms Goering, E., Lafkioti, M., Gold, S. Comment on "Spin and orbital magnetic moments of Fe3O4 {Physical Review Letters}, 96(3), 2006 (article)

[BibTex]

Effect of magnetic interaction on thermally activated reversal in a magnet with a distribution of energy barriers

Zhang, H. W., Zhang, S., Sun, J. R., Shen, B. G., Goll, D., Kronmüller, H.

{Journal of the Iron and Steel Research International}, 13(Suppl. 1):60-66, 2006 (article)

mms

[BibTex]

mms Zhang, H. W., Zhang, S., Sun, J. R., Shen, B. G., Goll, D., Kronmüller, H. Effect of magnetic interaction on thermally activated reversal in a magnet with a distribution of energy barriers {Journal of the Iron and Steel Research International}, 13(Suppl. 1):60-66, 2006 (article)

[BibTex]

A biomimetic climbing robot based on the gecko

Menon, C., Sitti, M.

Journal of Bionic Engineering, 3(3):115-125, 2006 (article)

[BibTex]

pi Menon, C., Sitti, M. A biomimetic climbing robot based on the gecko Journal of Bionic Engineering, 3(3):115-125, 2006 (article)

[BibTex]

Proximal probes based nanorobotic drawing of polymer micro/nanofibers

Nain, A. S., Amon, C., Sitti, M.

IEEE transactions on nanotechnology, 5(5):499-510, IEEE, 2006 (article)

[BibTex]

pi Nain, A. S., Amon, C., Sitti, M. Proximal probes based nanorobotic drawing of polymer micro/nanofibers IEEE transactions on nanotechnology, 5(5):499-510, IEEE, 2006 (article)

[BibTex]

Rocking Stamper and Jumping Snake from a Dynamical System Approach to Artificial Life

Der, R., Hesse, F., Martius, G.

Adaptive Behavior, 14(2):105-115, 2006 (article)

Abstract

Dynamical systems offer intriguing possibilities as a substrate for the generation of behavior because of their rich behavioral complexity. However this complexity together with the largely covert relation between the parameters and the behavior of the agent is also the main hindrance in the goal-oriented design of a behavior system. This paper presents a general approach to the self-regulation of dynamical systems so that the design problem is circumvented. We consider the controller (a neural net work) as the mediator for changes in the sensor values over time and define a dynamics for the parameters of the controller by maximizing the dynamical complexity of the sensorimotor loop under the condition that the consequences of the actions taken are still predictable. This very general principle is given a concrete mathematical formulation and is implemented in an extremely robust and versatile algorithm for the parameter dynamics of the controller. We consider two different applications, a mechanical device called the rocking stamper and the ODE simulations of a "snake" with five degrees of freedom. In these and many other examples studied we observed various behavior modes of high dynamical complexity.

DOI [BibTex]

al Der, R., Hesse, F., Martius, G. Rocking Stamper and Jumping Snake from a Dynamical System Approach to Artificial Life Adaptive Behavior, 14(2):105-115, 2006 (article)

DOI [BibTex]

Magnetic vortex core reversal by excitation with short bursts of an alternating field

Van Waeyenberge, B., Puzic, A., Stoll, H., Chou, K. W., Tyliszczak, T., Hertel, R., Fähnle, M., Bruckl, H., Rott, K., Reiss, G., Neudecker, I., Weiss, D., Back, C. H., Schütz, G.

{Nature}, 444(7118):461-464, 2006 (article)

mms

[BibTex]

mms Van Waeyenberge, B., Puzic, A., Stoll, H., Chou, K. W., Tyliszczak, T., Hertel, R., Fähnle, M., Bruckl, H., Rott, K., Reiss, G., Neudecker, I., Weiss, D., Back, C. H., Schütz, G. Magnetic vortex core reversal by excitation with short bursts of an alternating field {Nature}, 444(7118):461-464, 2006 (article)

[BibTex]

Structure and magnetism in bcc-based iron-cobalt alloys

Diaz-Ortiz, A., Drautz, R., Fähnle, M., Dosch, H., Sanchez, J. M.

{Physical Review B}, 73, 2006 (article)

mms

[BibTex]

mms Diaz-Ortiz, A., Drautz, R., Fähnle, M., Dosch, H., Sanchez, J. M. Structure and magnetism in bcc-based iron-cobalt alloys {Physical Review B}, 73, 2006 (article)

[BibTex]

Role of nonequilibrium conduction electrons on the magnetization dynamics of ferromagnets in the s-d model

Fähnle, M., Singer, R., Steiauf, D., Antropov, V. P.

{Physical Review B}, 73, 2006 (article)

mms

[BibTex]

mms Fähnle, M., Singer, R., Steiauf, D., Antropov, V. P. Role of nonequilibrium conduction electrons on the magnetization dynamics of ferromagnets in the s-d model {Physical Review B}, 73, 2006 (article)

[BibTex]

Hydrogen adsorption in metal-organic frameworks: Cu-MOFs and Zn-MOFs compared

Panella, B., Hirscher, M., Pütter, H., Müller, U.

{Advanced Functional Materials}, 16, pages: 520-524, 2006 (article)

mms

[BibTex]

mms Panella, B., Hirscher, M., Pütter, H., Müller, U. Hydrogen adsorption in metal-organic frameworks: Cu-MOFs and Zn-MOFs compared {Advanced Functional Materials}, 16, pages: 520-524, 2006 (article)

[BibTex]

Hardness of nanostructured Al-Zn, Al-Mg and Al-Zn-Mg alloys obtained by high-pressure torsion

Mazilkin, A.A., Baretzky, B., Enders, S., Kogtenkova, O.A., Straumal, B.B., Rabkin, E., Valiev, R.Z.

{Defect and Diffusion Forum}, 249, pages: 155-160, 2006 (article)

mms

[BibTex]

mms Mazilkin, A.A., Baretzky, B., Enders, S., Kogtenkova, O.A., Straumal, B.B., Rabkin, E., Valiev, R.Z. Hardness of nanostructured Al-Zn, Al-Mg and Al-Zn-Mg alloys obtained by high-pressure torsion {Defect and Diffusion Forum}, 249, pages: 155-160, 2006 (article)

[BibTex]

2005

Kernel Methods for Measuring Independence

Gretton, A., Herbrich, R., Smola, A., Bousquet, O., Schölkopf, B.

Journal of Machine Learning Research, 6, pages: 2075-2129, December 2005 (article)

Abstract

We introduce two new functionals, the constrained covariance and the kernel mutual information, to measure the degree of independence of random variables. These quantities are both based on the covariance between functions of the random variables in reproducing kernel Hilbert spaces (RKHSs). We prove that when the RKHSs are universal, both functionals are zero if and only if the random variables are pairwise independent. We also show that the kernel mutual information is an upper bound near independence on the Parzen window estimate of the mutual information. Analogous results apply for two correlation-based dependence functionals introduced earlier: we show the kernel canonical correlation and the kernel generalised variance to be independence measures for universal kernels, and prove the latter to be an upper bound on the mutual information near independence. The performance of the kernel dependence functionals in measuring independence is verified in the context of independent component analysis.

PDF PostScript PDF [BibTex]

2005

ei Gretton, A., Herbrich, R., Smola, A., Bousquet, O., Schölkopf, B. Kernel Methods for Measuring Independence Journal of Machine Learning Research, 6, pages: 2075-2129, December 2005 (article)

PDF PostScript PDF [BibTex]

A Unifying View of Sparse Approximate Gaussian Process Regression

Quinonero Candela, J., Rasmussen, C.

Journal of Machine Learning Research, 6, pages: 1935-1959, December 2005 (article)

Abstract

We provide a new unifying view, including all existing proper probabilistic sparse approximations for Gaussian process regression. Our approach relies on expressing the effective prior which the methods are using. This allows new insights to be gained, and highlights the relationship between existing methods. It also allows for a clear theoretically justified ranking of the closeness of the known approximations to the corresponding full GPs. Finally we point directly to designs of new better sparse approximations, combining the best of the existing strategies, within attractive computational constraints.

PDF [BibTex]

ei Quinonero Candela, J., Rasmussen, C. A Unifying View of Sparse Approximate Gaussian Process Regression Journal of Machine Learning Research, 6, pages: 1935-1959, December 2005 (article)

PDF [BibTex]

Maximal Margin Classification for Metric Spaces

Hein, M., Bousquet, O., Schölkopf, B.

Journal of Computer and System Sciences, 71(3):333-359, October 2005 (article)

Abstract

In order to apply the maximum margin method in arbitrary metric spaces, we suggest to embed the metric space into a Banach or Hilbert space and to perform linear classification in this space. We propose several embeddings and recall that an isometric embedding in a Banach space is always possible while an isometric embedding in a Hilbert space is only possible for certain metric spaces. As a result, we obtain a general maximum margin classification algorithm for arbitrary metric spaces (whose solution is approximated by an algorithm of Graepel. Interestingly enough, the embedding approach, when applied to a metric which can be embedded into a Hilbert space, yields the SVM algorithm, which emphasizes the fact that its solution depends on the metric and not on the kernel. Furthermore we give upper bounds of the capacity of the function classes corresponding to both embeddings in terms of Rademacher averages. Finally we compare the capacities of these function classes directly.

PDF PDF DOI [BibTex]

ei Hein, M., Bousquet, O., Schölkopf, B. Maximal Margin Classification for Metric Spaces Journal of Computer and System Sciences, 71(3):333-359, October 2005 (article)

PDF PDF DOI [BibTex]

Selective integration of multiple biological data for supervised network inference

Kato, T., Tsuda, K., Asai, K.

Bioinformatics, 21(10):2488 , October 2005 (article)

PDF [BibTex]

ei Kato, T., Tsuda, K., Asai, K. Selective integration of multiple biological data for supervised network inference Bioinformatics, 21(10):2488 , October 2005 (article)

PDF [BibTex]

Assessing Approximate Inference for Binary Gaussian Process Classification

Kuss, M., Rasmussen, C.

Journal of Machine Learning Research, 6, pages: 1679 , October 2005 (article)

Abstract

Gaussian process priors can be used to define flexible, probabilistic classification models. Unfortunately exact Bayesian inference is analytically intractable and various approximation techniques have been proposed. In this work we review and compare Laplace‘s method and Expectation Propagation for approximate Bayesian inference in the binary Gaussian process classification model. We present a comprehensive comparison of the approximations, their predictive performance and marginal likelihood estimates to results obtained by MCMC sampling. We explain theoretically and corroborate empirically the advantages of Expectation Propagation compared to Laplace‘s method.

PDF PDF [BibTex]

ei Kuss, M., Rasmussen, C. Assessing Approximate Inference for Binary Gaussian Process Classification Journal of Machine Learning Research, 6, pages: 1679 , October 2005 (article)

PDF PDF [BibTex]

Clustering on the Unit Hypersphere using von Mises-Fisher Distributions

Banerjee, A., Dhillon, I., Ghosh, J., Sra, S.

Journal of Machine Learning Research, 6, pages: 1345-1382, September 2005 (article)

Abstract

Several large scale data mining applications, such as text categorization and gene expression analysis, involve high-dimensional data that is also inherently directional in nature. Often such data is L2 normalized so that it lies on the surface of a unit hypersphere. Popular models such as (mixtures of) multi-variate Gaussians are inadequate for characterizing such data. This paper proposes a generative mixture-model approach to clustering directional data based on the von Mises-Fisher (vMF) distribution, which arises naturally for data distributed on the unit hypersphere. In particular, we derive and analyze two variants of the Expectation Maximization (EM) framework for estimating the mean and concentration parameters of this mixture. Numerical estimation of the concentration parameters is non-trivial in high dimensions since it involves functional inversion of ratios of Bessel functions. We also formulate two clustering algorithms corresponding to the variants of EM that we derive. Our approach provides a theoretical basis for the use of cosine similarity that has been widely employed by the information retrieval community, and obtains the spherical kmeans algorithm (kmeans with cosine similarity) as a special case of both variants. Empirical results on clustering of high-dimensional text and gene-expression data based on a mixture of vMF distributions show that the ability to estimate the concentration parameter for each vMF component, which is not present in existing approaches, yields superior results, especially for difficult clustering tasks in high-dimensional spaces.

PDF [BibTex]

ei Banerjee, A., Dhillon, I., Ghosh, J., Sra, S. Clustering on the Unit Hypersphere using von Mises-Fisher Distributions Journal of Machine Learning Research, 6, pages: 1345-1382, September 2005 (article)

PDF [BibTex]

Support Vector Machines for 3D Shape Processing

Steinke, F., Schölkopf, B., Blanz, V.

Computer Graphics Forum, 24(3, EUROGRAPHICS 2005):285-294, September 2005 (article)

Abstract

We propose statistical learning methods for approximating implicit surfaces and computing dense 3D deformation fields. Our approach is based on Support Vector (SV) Machines, which are state of the art in machine learning. It is straightforward to implement and computationally competitive; its parameters can be automatically set using standard machine learning methods. The surface approximation is based on a modified Support Vector regression. We present applications to 3D head reconstruction, including automatic removal of outliers and hole filling. In a second step, we build on our SV representation to compute dense 3D deformation fields between two objects. The fields are computed using a generalized SVMachine enforcing correspondence between the previously learned implicit SV object representations, as well as correspondences between feature points if such points are available. We apply the method to the morphing of 3D heads and other objects.

PDF [BibTex]

ei Steinke, F., Schölkopf, B., Blanz, V. Support Vector Machines for 3D Shape Processing Computer Graphics Forum, 24(3, EUROGRAPHICS 2005):285-294, September 2005 (article)

PDF [BibTex]

Fast Protein Classification with Multiple Networks

Tsuda, K., Shin, H., Schölkopf, B.

Bioinformatics, 21(Suppl. 2):59-65, September 2005 (article)

Abstract

Support vector machines (SVM) have been successfully used to classify proteins into functional categories. Recently, to integrate multiple data sources, a semidefinite programming (SDP) based SVM method was introduced Lanckriet et al (2004). In SDP/SVM, multiple kernel matrices corresponding to each of data sources are combined with weights obtained by solving an SDP. However, when trying to apply SDP/SVM to large problems, the computational cost can become prohibitive, since both converting the data to a kernel matrix for the SVM and solving the SDP are time and memory demanding. Another application-specific drawback arises when some of the data sources are protein networks. A common method of converting the network to a kernel matrix is the diffusion kernel method, which has time complexity of O(n^3), and produces a dense matrix of size n x n. We propose an efficient method of protein classification using multiple protein networks. Available protein networks, such as a physical interaction network or a metabolic network, can be directly incorporated. Vectorial data can also be incorporated after conversion into a network by means of neighbor point connection. Similarly to the SDP/SVM method, the combination weights are obtained by convex optimization. Due to the sparsity of network edges, the computation time is nearly linear in the number of edges of the combined network. Additionally, the combination weights provide information useful for discarding noisy or irrelevant networks. Experiments on function prediction of 3588 yeast proteins show promising results: the computation time is enormously reduced, while the accuracy is still comparable to the SDP/SVM method.

PDF Web DOI [BibTex]

ei Tsuda, K., Shin, H., Schölkopf, B. Fast Protein Classification with Multiple Networks Bioinformatics, 21(Suppl. 2):59-65, September 2005 (article)

PDF Web DOI [BibTex]

Iterative Kernel Principal Component Analysis for Image Modeling

Kim, K., Franz, M., Schölkopf, B.

IEEE Transactions on Pattern Analysis and Machine Intelligence, 27(9):1351-1366, September 2005 (article)

Abstract

In recent years, Kernel Principal Component Analysis (KPCA) has been suggested for various image processing tasks requiring an image model such as, e.g., denoising or compression. The original form of KPCA, however, can be only applied to strongly restricted image classes due to the limited number of training examples that can be processed. We therefore propose a new iterative method for performing KPCA, the Kernel Hebbian Algorithm which iteratively estimates the Kernel Principal Components with only linear order memory complexity. In our experiments, we compute models for complex image classes such as faces and natural images which require a large number of training examples. The resulting image models are tested in single-frame super-resolution and denoising applications. The KPCA model is not specifically tailored to these tasks; in fact, the same model can be used in super-resolution with variable input resolution, or denoising with unknown noise characteristics. In spite of this, both super-resolution a nd denoising performance are comparable to existing methods.

Web DOI [BibTex]

ei Kim, K., Franz, M., Schölkopf, B. Iterative Kernel Principal Component Analysis for Image Modeling IEEE Transactions on Pattern Analysis and Machine Intelligence, 27(9):1351-1366, September 2005 (article)

Web DOI [BibTex]

Phenotypic characterization of chondrosarcoma-derived cell lines

Schorle, C., Finger, F., Zien, A., Block, J., Gebhard, P., Aigner, T.

Cancer Letters, 226(2):143-154, August 2005 (article)

Abstract

Gene expression profiling of three chondrosarcoma derived cell lines (AD, SM, 105KC) showed an increased proliferative activity and a reduced expression of chondrocytic-typical matrix products compared to primary chondrocytes. The incapability to maintain an adequate matrix synthesis as well as a notable proliferative activity at the same time is comparable to neoplastic chondrosarcoma cells in vivo which cease largely cartilage matrix formation as soon as their proliferative activity increases. Thus, the investigated cell lines are of limited value as substitute of primary chondrocytes but might have a much higher potential to investigate the behavior of neoplastic chondrocytes, i.e. chondrosarcoma biology.

Web [BibTex]

ei Schorle, C., Finger, F., Zien, A., Block, J., Gebhard, P., Aigner, T. Phenotypic characterization of chondrosarcoma-derived cell lines Cancer Letters, 226(2):143-154, August 2005 (article)

Web [BibTex]

Local Rademacher Complexities

Bartlett, P., Bousquet, O., Mendelson, S.

The Annals of Statistics, 33(4):1497-1537, August 2005 (article)

Abstract

We propose new bounds on the error of learning algorithms in terms of a data-dependent notion of complexity. The estimates we establish give optimal rates and are based on a local and empirical version of Rademacher averages, in the sense that the Rademacher averages are computed from the data, on a subset of functions with small empirical error. We present some applications to classification and prediction with convex function classes, and with kernel classes in particular.

PDF PostScript Web [BibTex]

ei Bartlett, P., Bousquet, O., Mendelson, S. Local Rademacher Complexities The Annals of Statistics, 33(4):1497-1537, August 2005 (article)

PDF PostScript Web [BibTex]

Learning the Kernel with Hyperkernels

Ong, CS., Smola, A., Williamson, R.

Journal of Machine Learning Research, 6, pages: 1043-1071, July 2005 (article)

Abstract

This paper addresses the problem of choosing a kernel suitable for estimation with a Support Vector Machine, hence further automating machine learning. This goal is achieved by defining a Reproducing Kernel Hilbert Space on the space of kernels itself. Such a formulation leads to a statistical estimation problem similar to the problem of minimizing a regularized risk functional. We state the equivalent representer theorem for the choice of kernels and present a semidefinite programming formulation of the resulting optimization problem. Several recipes for constructing hyperkernels are provided, as well as the details of common machine learning problems. Experimental results for classification, regression and novelty detection on UCI data show the feasibility of our approach.

PDF [BibTex]

ei Ong, CS., Smola, A., Williamson, R. Learning the Kernel with Hyperkernels Journal of Machine Learning Research, 6, pages: 1043-1071, July 2005 (article)

PDF [BibTex]

Image Reconstruction by Linear Programming

Tsuda, K., Rätsch, G.

IEEE Transactions on Image Processing, 14(6):737-744, June 2005 (article)

Abstract

One way of image denoising is to project a noisy image to the subspace of admissible images derived, for instance, by PCA. However, a major drawback of this method is that all pixels are updated by the projection, even when only a few pixels are corrupted by noise or occlusion. We propose a new method to identify the noisy pixels by l1-norm penalization and to update the identified pixels only. The identification and updating of noisy pixels are formulated as one linear program which can be efficiently solved. In particular, one can apply the upsilon trick to directly specify the fraction of pixels to be reconstructed. Moreover, we extend the linear program to be able to exploit prior knowledge that occlusions often appear in contiguous blocks (e.g., sunglasses on faces). The basic idea is to penalize boundary points and interior points of the occluded area differently. We are also able to show the upsilon property for this extended LP leading to a method which is easy to use. Experimental results demonstrate the power of our approach.

PDF DOI [BibTex]