CoRE Kernels
This work addresses the challenge of efficiently handling non-binary sparse data in machine learning, though it appears incremental as it builds on existing kernel methods with hashing adaptations.
The authors tackled the problem of high-dimensional, sparse, non-binary data in applications like vision by proposing CoRE kernels, which are simple and parameter-free, and demonstrated their effectiveness in a classification experiment. They also developed probabilistic hashing algorithms to transform these nonlinear kernels into linear ones, making them more practical for large-scale industrial use.
The term "CoRE kernel" stands for correlation-resemblance kernel. In many applications (e.g., vision), the data are often high-dimensional, sparse, and non-binary. We propose two types of (nonlinear) CoRE kernels for non-binary sparse data and demonstrate the effectiveness of the new kernels through a classification experiment. CoRE kernels are simple with no tuning parameters. However, training nonlinear kernel SVM can be (very) costly in time and memory and may not be suitable for truly large-scale industrial applications (e.g. search). In order to make the proposed CoRE kernels more practical, we develop basic probabilistic hashing algorithms which transform nonlinear kernels into linear kernels.