Efficient and accurate ℓ p-norm multiple kernel learning

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Authors

  • Marius Kloft
  • Ulf Brefeld
  • Soren Sonnenburg
  • Pavel Laskov
  • Klaus Robert Müller
  • Alexander Zien

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. Unfortunately, ℓ 1-norm MKL is hardly observed to outperform trivial baselines in practical applications. To allow for robust kernel mixtures, we generalize MKL to arbitrary ℓ p-norms. We devise new insights on the connection between several existing MKL formulations and develop two efficient interleaved optimization strategies for arbitrary p > 1. Empirically, we demonstrate that the interleaved optimization strategies are much faster compared to the traditionally used wrapper approaches. Finally, we apply ℓp-norm MKL to real-world problems from computational biology, showing that non-sparse MKL achieves accuracies that go beyond the state-of-the-art.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems 22 : Proceedings of the 23rd Annual Conference on Neural Information Processing Systems 2009
EditorsYoshua Bengio, Dale Schuurmans, John Lafferty, Chris Williams, Aron Culotta
Number of pages9
PublisherNeural Information Processing Systems
Publication date2009
Pages997-1005
ISBN (print)978-161567911-9
Publication statusPublished - 2009
Externally publishedYes
Event23rd Annual Conference on Neural Information Processing Systems, NIPS 2009 - Hyatt Regency Vancouver, Vancouver, BC, Canada
Duration: 07.12.200910.12.2009
Conference number: 23
https://nips.cc/Conferences/2009

Recently viewed

Publications

  1. The Influence of Note-taking on Mathematical Solution Processes while Working on Reality-Based Tasks
  2. Using protochirons for three-dimensional coding of certain chemical structures.
  3. On finding nonisomorphic connected subgraphs and distinct molecular substructures.
  4. Improving students’ science text comprehension through metacognitive self-regulation when applying learning strategies
  5. Dynamic environment modelling and prediction for autonomous systems
  6. Towards improved dispatching rules for complex shop floor scenarios - A genetic programming approach
  7. Data-Driven flood detection using neural networks
  8. Using augmented video to test in-car user experiences of context analog HUDs
  9. Data-driven and physics-based modelling of process behaviour and deposit geometry for friction surfacing
  10. Essentializing the binary self
  11. Overcoming Multi-legacy Application Challenges through Building Dynamic Capabilities for Low-Code Adoption
  12. Integration of Environmental Management Information Systems and ERP systems using Integration Platforms
  13. XOperator - Interconnecting the semantic web and instant messaging networks
  14. Reality-Based Tasks with Complex-Situations
  15. A geometric approach for controlling an electromagnetic actuator with the help of a linear Model Predictive Control
  16. Simple saturated PID control for fast transient of motion systems
  17. Faulty Process Detection Using Machine Learning Techniques
  18. Supporting the Development and Implementation of a Digitalization Strategy in SMEs through a Lightweight Architecture-based Method
  19. Identification of conductive fiber parameters with transcutaneous electrical nerve stimulation signal using RLS algorithm
  20. HAWK - hybrid question answering using linked data
  21. Knowledge-Enhanced Language Models Are Not Bias-Proof
  22. Advances in Dynamics, Optimization and Computation
  23. A general structural property in wavelet packets for detecting oscillation and noise components in signal analysis
  24. The effects of different on-line adaptive response time limits on speed and amount of learning in computer assisted instruction and intelligent tutoring
  25. Is too much help an obstacle? Effects of interactivity and cognitive style on learning with dynamic versus non-dynamic visualizations with narrative explanations
  26. Neural relational inference for disaster multimedia retrieval
  27. Mechanical performance prediction for friction riveting joints of dissimilar materials via machine learning
  28. The impact of goal focus, task type and group size on synchronous net-based collaborative learning discourses
  29. Spaces with a temper
  30. Need Satisfaction and Optimal Functioning at Leisure and Work: A Longitudinal Validation Study of the DRAMMA Model
  31. A MODEL FOR QUANTIFICATION OF SOFTWARE COMPLEXITY
  32. Recent Advances in Intelligent Algorithms for Fault Detection and Diagnosis