l p -Norm Multiple Kernel Learning

By Marius Kloft, Ulf Brefeld, Sören Sonnenburg, Alexander Zien. 2011

356 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Molecular Biology

Citations in the corpus

Modeling Spatial-Temporal Clues in a Hybrid Deep Learning Framework for Video Classification (2015, methodology): “rate SVM classiers are trained for each video-level CNN feature and the prediction outputs are averaged; (6) Multiple Kernel Learning (SVM-MKL), which combines the two features with the ‘ p-norm MKL [20] by xing p = 2; (7) Early Fusion with Neural Networks (NN-EF), which concatenates the two feature”
Scalable Matrix-valued Kernel Learning for High-dimensional Nonlinear Multivariate Regression and Granger Causality (2012, methodology): “This complements and extends generalization results in the multiple kernel learning literature for the scalar case [11, 38, 19, 24, 20].”
RNA-protein binding motifs mining with a new hybrid deep learning based cross-domain knowledge integration approach (2016, background): “The other implementation of feature-level fusion is the multiple-kernel learning, which design multiple kernels for different features, and then combine them together [14, 15].”
Multi-Task Multiple Kernel Relationship Learning (2016, background): “irically performs better than relations among the original feature spaces. There have been a few attempts to imposing higherorder relationship between kernel spaces using kernel weights. Kloft et. al [12] propose a non-isotropic norm such as q β⊤Σ−1β on kernels weights β to induce the relationship b”
A Binary Classification Framework for Two-Stage Multiple Kernel Learning (2012, methodology): “This one-stage approach was first proposed by (Lanckriet et al., 2004) and has since received significant attention (Rakotomamonjy et al., 2007; Sonnenburg et al., 2006; Cortes et al., 2010a; Kloft et al., 2011; Bach, 2008; Zien & Ong, 2007; Cortes et al., 2009; Sindhwani & Lozano, 2011).”
How to Scale Up Kernel Methods to Be As Good As Deep Neural Nets (2014, background): “kernel function to use implicitly 5 species the basis functions. But then how to select the best kernel function? One popular paradigm to address the latter problem is multiple kernel learning (MKL) [27, 1, 12, 25]. That is, starting from a collection of base kernels, the algorithm identies the best”
Predicting MOOC Dropout over Weeks Using Machine Learning Methods (2014, methodology): “While this paper focuses on clickstream data, the approach could in principle also combined with forum data (e.g., using multiple kernel learning (Kloft et al., 2011)), which we would like to tackle in future work.”
Multiple Operator-valued Kernel Learning (2012, methodology): “Hence, we propose to borrow from [15], and consider a block-coordinate descent method.”
Exploiting Feature and Class Relationships in Video Categorization with Regularized Deep Neural Networks (2015, methodology): “Fusion with SVM (SVM-LF). An SVM classiﬁer is trained for each feature and prediction results are then combined. 6) Multiple Kernel Learning (SVM-MKL). We perform feature fusion with the ‘ p-Norm MKL [59] by ﬁxing p= 2. MKL is able to learn dynamic fusion weights. For the above EF/LF approaches 1–4,”
Learning With Augmented Features for Supervised and Semi-Supervised Heterogeneous Domain Adaptation (2014, methodology): “In this subsection, we show that (7) can be equivalently reformulated as a convex MKL problem so that the global solution can be guaranteed by using the existing MKL solvers [19].”