Abstract:Multiple kernel learning (MKL) is demonstrated to be flexible and effective in depicting heterogeneous data sources since MKL can introduce multiple kernels rather than a single fixed kernel into applications. However, MKL would get a high time and space complexity in contrast to single kernel learning, which is not expected in real-world applications. Meanwhile, it is known that the kernel mapping ways of MKL generally have two forms including implicit kernel mapping and empirical kernel mapping (EKM), where the latter is less attracted. In this paper, we focus on the MKL with the EKM, and propose a reduced multiple empirical kernel learning machine named RMEKLM for short. To the best of our knowledge, it is the first to reduce both time and space complexity of the MKL with EKM. Different from the existing MKL, the proposed RMEKLM adopts the Gauss Elimination technique to extract a set of feature vectors, which is validated that doing so does not lose much information of the original feature space. Then RMEKLM adopts the extracted feature vectors to span a reduced orthonormal subspace of the feature space, which is visualized in terms of the geometry structure. It can be demonstrated that the spanned subspace is isomorphic to the original feature space, which means that the dot product of two vectors in the original feature space is equal to that of the two corresponding vectors in the generated orthonormal subspace. More importantly, the proposed RMEKLM brings a simpler computation and meanwhile needs a less storage space, especially in the processing of testing. Finally, the experimental results show that RMEKLM owns a much efficient and effective performance in terms of both complexity and classification. The contributions of this paper can be given as follows: (1) by mapping the input space into an orthonormal subspace, the geometry of the generated subspace is visualized; (2) this paper first reduces both the time and space complexity of the EKM-based MKL; (3) this paper adopts the Gauss Elimination, one of the on-the-shelf techniques, to generate a basis of the original feature space, which is stable and efficient.

Multiple Universum Empirical Kernel Learning

Multi-view Learning with Universum

Multiple Kernel Subspace Learning for Clustering and Classification

Structural multiple empirical kernel learning

Multiple Empirical Kernel Learning with Discriminant Locality Preservation

Unsupervised Multiple Kernel Learning

Multiple Empirical Kernel Learning with Locality Preserving Constraint

Multiple empirical kernel learning based on local information

Semi‐supervised multiple empirical kernel learning with pseudo empirical loss and similarity regularization

MKEL: Multiple Kernel Ensemble Learning via Unified Ensemble Loss for Image Classification

Reduced Multiple Empirical Kernel Learning Machine

A new multi-task learning method with universum data

Double-fold localized multiple matrix learning machine with Universum

MREKLM: A Fast Multiple Empirical Kernel Learning Machine

Locality Density-Based Fuzzy Multiple Empirical Kernel Learning

Multi-task twin support vector machine with Universum data

Multiple Empirical Kernel Learning with Majority Projection for Imbalanced Problems.

MPEKDyL: Efficient Multi-Partial Empirical Kernel Dynamic Learning

Coupled Multiple Kernel Learning for Supervised Classification.

Creating Universum for class imbalance via locality and its application in multiview subspace learning

Multi-view Dimensionality Reduction Based on Universum Learning