This paper presents gradKCCA, a large-scale sparse non-linear canonical correlation method. Like Kernel Canonical Correlation Analysis (KCCA), our method finds non-linear relations through kernel functions, but it does not rely on a kernel matrix, a known bottleneck for scaling up kernel methods. gradKCCA corresponds to solving KCCA with the additional constraint that the canonical projection directions in the kernel-induced feature space have preimages in the original data space. Firstly, this modification allows us to very efficiently maximize kernel canonical correlation through an alternating projected gradient algorithm working in the original data space. Secondly, we can control the sparsity of the projection directions by constraining the $\ell_1$ norm of the preimages of the projection directions, facilitating the interpretation of the discovered patterns, which is not available through KCCA. Our empirical experiments demonstrate that gradKCCA outperforms state-of-the-art CCA methods in terms of speed and robustness to noise both in simulated and real-world datasets.

We present SCCA-HSIC, a method for finding sparse non-linear multivariate relations in high-dimensional settings by maximizing the Hilbert-Schmidt Independence Criterion (HSIC). We propose efficient optimization algorithms using a projected stochastic gradient and Nyström approximation of HSIC. We demonstrate the favourable performance of SCCA-HSIC over competing methods in detecting multivariate non-linear relations both in simulation studies, with varying numbers of related variables, noise variables, and samples, as well as in real datasets.