【摘 要】
：
Effective and efficient treatment of massive data sets has become increasing important in this age of information explosion.Most machine learning and data a
【机 构】
：
Univ.of California Berkeley
论文部分内容阅读
Effective and efficient treatment of massive data sets has become increasing important in this age of information explosion.Most machine learning and data analysis algorithms for massive data sets require huge amounts of computational time.
其他文献
In this talk,I present a new approximate dynamic programming(ADP)approach to solve a category of resource-constrained project scheduling problems.
Scheduling theory has received a wide coverage in the literature on operations research and discrete optimization over the last five decades or so,but the l
We study the integrable discretization of the defocusing nonlinear Schrödinger equation,namely the Ablowitz-Ladik model.The asymptotic behavior of the solu
We propose a scheme of creating a tunable highly nonlinear defect in a one-dimensional photonic crystal.The defect consists of an atomic cell filled in with
In a few recent years there was growing interest in propagation of nonlinear waves in media with gain and losses.These are systems with the parity-time(PT)s
In recent years,much attention has been paid to the study of slowlight and light memory in various physical systems.In this talk,I shall report our recent r
Given a matrix subspace,there are relevant reasons to ask for a basis of lowest rank.A greedy algorithm will provably find it,but the subproblems are NP-har
We introduce randomized iterative algorithms inspired by the diffusion Monte Carlo algorithm for some common tasks in numerical linear algebra.
There are algorithms to simultaneously reduce a pair of symmetric matrices to tridiagonal-tridiagonal form with a congruent transformation,but these algorit
We propose a dynamic approach to sparse recovery under noisy linear measurements,which solve a dilemma in statistics discovered by Fan and Li in 2001: LASSO