Stopping rules for stochastic approximation

2014 Fiscal Year Final Research Report

• Project/Area Number: 24760344
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Control engineering
• Research Institution: Osaka University
• Principal Investigator: WADA Takayuki 大阪大学, 情報科学研究科, 助教 (60599207)
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 制御工学 / 数理工学 / アルゴリズム / 確率近似法

Outline of Final Research Achievements

We have investigated stopping rules of stochastic approximation, which is a recursive procedure to find the solution to an unknown nonlinear equation via noisy measurements. In fact, we have shown that there is a high probability that the distance between the exact solution and the candidate solution is less than a specified tolerance level when the stochastic approximation stops according to our stopping rule. From our stopping rules, we can know the necessary number of recursions in advance. Furthermore, since the number of recursions required by the stopping rule is a polynomial function of the problem size, this result also provides detailed analysis of the computational complexity of stochastic approximation.

Free Research Field

制御工学