20.500.12556/RUL-125566 Approximate multiple kernel learning with least-angle regression Kernel methods provide a principled way for general data representations. Multiple kernel learning and kernel approximation are often treated as separate tasks, with considerable savings in time and memory expected if the two are performed simultaneously. Our proposed Mklaren algorithm selectively approximates multiple kernel matrices in regression. It uses Incomplete Cholesky Decomposition and Least-angle regression (LAR) to select basis functions, achieving linear complexity both in the number of data points and kernels. Since it approximates kernel matrices rather than functions, it allows to combine an arbitrary set of kernels. Compared to single kernel-based approximations, it selectively approximates different kernels in different regions of the input spaces. The LAR criterion provides a robust selection of inducing points in noisy settings, and an accurate modelling of regression functions in continuous and discrete input spaces. Among general kernel matrix decompositions, Mklaren achieves minimal approximation rank required for performance comparable to using the exact kernel matrix, at a cost lower than 1% of required operations. Finally, we demonstrate the scalability and interpretability in settings with millions of data points and thousands of kernels. kernel methods kernel approximation multiple kernel learning least-angle regression jedrne metode aproksimacija jeder učenje z več jedrnimi funkcijami regresija najmanjših kotov true false true Angleški jezik Slovenski jezik Članek v reviji 2021-03-25 14:27:46 2021-03-25 14:27:47 2023-03-22 13:42:28 0000-00-00 00:00:00 2019 0 0 Str. 245-258 Vol. 340 May 2019 0000-00-00 Zaloznikova Objavljeno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004 0925-2312 10.1016/j.neucom.2019.02.030 1538162883 172315 RAZ_Strazar_Martin_2019.pdf RAZ_Strazar_Martin_2019.pdf 1 AB5D1179451FAEB7E5E0E1BE7750DC4A bbc30d8b6abc303ad62f19ed4ef043a8fda06dc6ff41972e2e6d66bbb4e4ee60 bc14f140-a1bb-11eb-a523-00155dcfd717 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=141328 https://www.sciencedirect.com/science/article/pii/S0925231219302449 1 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=141327 Fakulteta za računalništvo in informatiko 0 0 0