p-norm multiple kernel learning

Research output: Journal contributionsJournal articlesResearchpeer-review

Authors

Learning linear combinations of multiple kernels is an appealing strategy when the right choice of features is unknown. Previous approaches to multiple kernel learning (MKL) promote sparse kernel combinations to support interpretability and scalability. Unfortunately, this ℓ1norm MKL is rarely observed to outperform trivial baselines in practical applications. To allow for robust kernel mixtures that generalize well, we extend MKL to arbitrary norms. We devise new insights on the connection between several existing MKL formulations and develop two efficient interleaved optimization strategies for arbitrary norms, that is ℓp -norms with p ≥ 1. This interleaved optimization is much faster than the commonly used wrapper approaches, as demonstrated on several data sets. A theoretical analysis and an experiment on controlled artificial data shed light on the appropriateness of sparse, non-sparse and ℓ-norm MKL in various scenarios. Importantly, empirical applications of ℓp-norm MKL to three real-world problems from computational biology show that non-sparse MKL achieves accuracies that surpass the state-of-the-art. Data sets, source code to reproduce the experiments, implementations of the algorithms, and further information are available at http://doc.ml.tu-berlin.de/nonsparse-mkl/.

Original languageEnglish
JournalJournal of Machine Learning Research
Volume12
Pages (from-to)953-997
Number of pages45
ISSN1532-4435
Publication statusPublished - 03.2011
Externally publishedYes

    Research areas

  • Bioinformatics, Block coordinate descent, Convex conjugate, Generalization bounds, Large scale optimization, Learning kernels, Multiple kernel learning, Non-sparse, Rademacher complexity, Support vector machine
  • Informatics

Recently viewed

Publications

  1. Individual Scans Fusion in Virtual Knowledge Base for Navigation of Mobile Robotic Group with 3D TVS
  2. Improve a 3D distance measurement accuracy in stereo vision systems using optimization methods’ approach
  3. Functional Richness and Relative Resilience of Bird Communities in Regions with Different Land Use Intensities
  4. Pressure fault recognition and compensation with an adaptive feedforward regulator in a controlled hybrid actuator within engine applications
  5. The impact of goal focus, task type and group size on synchronous net-based collaborative learning discourses
  6. Material flow analysis between dynamic modelling and life cycle assessment
  7. Active plasma resonance spectroscopy: Eigenfunction solutions in spherical geometry
  8. Collaborative benchmarking of functional-structural root architecture models
  9. Wavelet functions for rejecting spurious values
  10. Finite element modeling of laser beam welding for residual stress calculation
  11. Internet research differs from research on internet users
  12. Frame-based Data Factorizations
  13. Strengthening the transformative impulse while mainstreaming real-world labs: Lessons learned from three years of BaWü-Labs
  14. Gerbil – Benchmarking named entity recognition and linking consistently
  15. Introduction: Habitual Action, Automaticity, and Control
  16. Practice and carryover effects when using small interaction devices
  17. Teaching Sustainable Development in a Sensory and Artful Way — Concepts, Methods, and Examples
  18. Influence of Mg content in Al alloys on processing characteristics and dynamically recrystallized microstructure of friction surfacing deposits
  19. Stimulating Computing
  20. Comparison of three methods of length compensation in a parallel kinematic and their equivalence conditions
  21. Can a Revision of the Universal Service Scope Result in Substantive Change?
  22. Modeling and simulation of the heterogenous material behavior in thermal-sprayed coatings
  23. Sliding Mode Control of an Inductive Power Transmission System with Maximum Efficiency
  24. Short-arc measurement and fitting based on the bidirectional prediction of observed data
  25. Graph-Based Early-Fusion for Flood Detection
  26. Short and long-term dominance of negative information in shaping public energy perceptions
  27. Deconstructing and reconstructing diversity in client-provider-relationships of social work
  28. A New Approach for Optimal Solving Cyclic and Non-Cyclic Bus Drvier Rostering Problems
  29. Vielfalt des Alterns - Differenz oder Integration?