Scalable Pairwise Kernel Learning with Stochastic Vec Trick
Researchers have introduced SPaiK, a novel kernel learning method designed for pairwise settings that significantly reduces computational and memory demands. The core innovation is the stochastic generalized vec trick (sGVT), an extension of the sparse Kronecker product multiplication algorithm, which facilitates efficient large-scale training with pairwise kernels. This advancement allows kernel-based pairwise learning to be applied to previously unmanageable dataset sizes, as demonstrated by evaluations on seven drug-target affinity datasets. AI
IMPACT Enables larger-scale applications of kernel-based pairwise learning, particularly in domains like drug-target affinity prediction.