A new stepsize rule in He and Zhou's alternating direction method
نویسندگان
چکیده
منابع مشابه
Online Alternating Direction Method
Online optimization has emerged as powerful tool in large scale optimization. In this paper, we introduce efficient online optimization algorithms based on the alternating direction method (ADM), which can solve online convex optimization under linear constraints where the objective could be non-smooth. We introduce new proof techniques for ADM in the batch setting, which yields a O(1/T ) conve...
متن کاملAn Alternating Direction Method
combination of a's yields the numerical coefficient for that combination. For example, in the summation for Su given in Table 2, 1/ao4 has as the ai*-1aB_*+i coefficient ai3a8 which, in turn, has the numerical coefficient 11. The ai2a3a6 combination has 2 ! X 1 ! X 1 ! as the product of the factorials of the exponents. If 11 X 3 ! X 1 ! is divided by 2 ! X 1 ! X 1 !, the result is 33 which is t...
متن کاملA New Alternating Direction Method for Linear Programming
It is well known that, for a linear program (LP) with constraint matrix A ∈ Rm×n, the Alternating Direction Method of Multiplier converges globally and linearly at a rate O((‖A‖F +mn) log(1/ )). However, such a rate is related to the problem dimension and the algorithm exhibits a slow and fluctuating “tail convergence” in practice. In this paper, we propose a new variable splitting method of LP...
متن کاملAn Alternating Direction Implicit Method for Modeling of Fluid Flow
This research includes of the numerical modeling of fluids in two-dimensional cavity. The cavity flow is an important theoretical problem. In this research, modeling was carried out based on an alternating direction implicit via Vorticity-Stream function formulation. It evaluates different Reynolds numbers and grid sizes. Therefore, for the flow field analysis and prove of the ability of the sc...
متن کاملStepsize Selection for Approximate Value Iteration and a New Optimal Stepsize Rule
Approximate value iteration is used in dynamic programming when we use random observations to estimate the value of being in a state. These observations are smoothed to approximate the expected value function, leading to the problem of choosing a stepsize (the weight given to the most recent observation). A stepsize of 1/n is a common (and provably convergent) choice. However, we prove that it ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Applied Mathematics Letters
سال: 2002
ISSN: 0893-9659
DOI: 10.1016/s0893-9659(01)00115-x