A Nonparametric Approach to Noisy and Costly
نویسندگان
چکیده
This paper describes Pairwise Bisection: a nonparametric approach to optimizing a noisy function with few function evaluations. The algorithm uses nonparametric reasoning about simple geometric relationships to nd minima eeciently. Two factors often frustrate optimization: noise and cost. Output can contain signiicant quantities of noise or error, while time or money allows for only a handful of experiments. Pairwise bisection is used here to attempt to automate the process of robust and eecient experiment design. Real world functions also tend to violate traditional assumptions of continuousness and Gaussian noise. Since nonparametric statistics do not depend on these assumptions, this algorithm can optimize a wide variety of phenomena with fewer restrictions placed on noise. The algorithm's performance is compared to that of three competing algorithms, Amoeba, PMAX, and Q2 on several diierent test functions. Results on these functions indicate competitive performance and superior resistance to noise.
منابع مشابه
A Nonparametric Approach to Noisy and Costly OptimizationBrigham
This paper describes Pairwise Bisection: a nonparametric approach to optimizing a noisy function with few function evaluations. The algorithm uses nonparametric reasoning about simple geometric relationships to nd minima eeciently. Two factors often frustrate optimization: noise and cost. Output can contain signiicant quantities of noise or error, while time or money allows for only a handful o...
متن کاملComparative Study of Particle Swarm Optimization and Genetic Algorithm Applied for Noisy Non-Linear Optimization Problems
Optimization of noisy non-linear problems plays a key role in engineering and design problems. These optimization problems can't be solved effectively by using conventional optimization methods. However, metaheuristic algorithms such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) seem very efficient to approach in these problems and became very popular. The efficiency of these ...
متن کاملStatistical Topology Using the Nonparametric Density Estimation and Bootstrap Algorithm
This paper presents approximate confidence intervals for each function of parameters in a Banach space based on a bootstrap algorithm. We apply kernel density approach to estimate the persistence landscape. In addition, we evaluate the quality distribution function estimator of random variables using integrated mean square error (IMSE). The results of simulation studies show a significant impro...
متن کاملRealized Volatility in Noisy Prices: a MSRV approach
Volatility is the primary measure of risk in modern finance and volatility estimation and inference has attracted substantial attention in the recent financial econometric literature, especially in high-frequency analyses. High-frequency prices carry a significant amount of noise. Therefore, there are two volatility components embedded in the returns constructed using high frequency prices: the...
متن کامل