Lock in feedback in sequential experiments

Maurits Kaptein, Davide Iannuzzi

Research output: Working paperOther research output

Abstract

We often encounter situations in which an experimenter wants to find, by sequential experimentation, $x_{max} = \arg\max_{x} f(x)$, where $f(x)$ is a (possibly unknown) function of a well controllable variable $x$. Taking inspiration from physics and engineering, we have designed a new method to address this problem. In this paper, we first introduce the method in continuous time, and then present two algorithms for use in sequential experiments. Through a series of simulation studies, we show that the method is effective for finding maxima of unknown functions by experimentation, even when the maximum of the functions drifts or when the signal to noise ratio is low.
Original languageEnglish
PublisherarXiv.org
Number of pages20
Publication statusPublished - 2 Feb 2015

Publication series

NamearXiv

Fingerprint

Experimentation
Experiment
Unknown
Continuous Time
Physics
Simulation Study
Engineering
Series

Keywords

  • cs.LG

Cite this

Kaptein, M., & Iannuzzi, D. (2015). Lock in feedback in sequential experiments. (arXiv). arXiv.org.
Kaptein, Maurits ; Iannuzzi, Davide. / Lock in feedback in sequential experiments. arXiv.org, 2015. (arXiv).
@techreport{7eda32f159ea4c17bae4e9b6a16f1731,
title = "Lock in feedback in sequential experiments",
abstract = "We often encounter situations in which an experimenter wants to find, by sequential experimentation, $x_{max} = \arg\max_{x} f(x)$, where $f(x)$ is a (possibly unknown) function of a well controllable variable $x$. Taking inspiration from physics and engineering, we have designed a new method to address this problem. In this paper, we first introduce the method in continuous time, and then present two algorithms for use in sequential experiments. Through a series of simulation studies, we show that the method is effective for finding maxima of unknown functions by experimentation, even when the maximum of the functions drifts or when the signal to noise ratio is low.",
keywords = "cs.LG",
author = "Maurits Kaptein and Davide Iannuzzi",
note = "20 Pages, 7 Figures",
year = "2015",
month = "2",
day = "2",
language = "English",
series = "arXiv",
publisher = "arXiv.org",
type = "WorkingPaper",
institution = "arXiv.org",

}

Kaptein, M & Iannuzzi, D 2015 'Lock in feedback in sequential experiments' arXiv, arXiv.org.

Lock in feedback in sequential experiments. / Kaptein, Maurits; Iannuzzi, Davide.

arXiv.org, 2015. (arXiv).

Research output: Working paperOther research output

TY - UNPB

T1 - Lock in feedback in sequential experiments

AU - Kaptein, Maurits

AU - Iannuzzi, Davide

N1 - 20 Pages, 7 Figures

PY - 2015/2/2

Y1 - 2015/2/2

N2 - We often encounter situations in which an experimenter wants to find, by sequential experimentation, $x_{max} = \arg\max_{x} f(x)$, where $f(x)$ is a (possibly unknown) function of a well controllable variable $x$. Taking inspiration from physics and engineering, we have designed a new method to address this problem. In this paper, we first introduce the method in continuous time, and then present two algorithms for use in sequential experiments. Through a series of simulation studies, we show that the method is effective for finding maxima of unknown functions by experimentation, even when the maximum of the functions drifts or when the signal to noise ratio is low.

AB - We often encounter situations in which an experimenter wants to find, by sequential experimentation, $x_{max} = \arg\max_{x} f(x)$, where $f(x)$ is a (possibly unknown) function of a well controllable variable $x$. Taking inspiration from physics and engineering, we have designed a new method to address this problem. In this paper, we first introduce the method in continuous time, and then present two algorithms for use in sequential experiments. Through a series of simulation studies, we show that the method is effective for finding maxima of unknown functions by experimentation, even when the maximum of the functions drifts or when the signal to noise ratio is low.

KW - cs.LG

M3 - Working paper

T3 - arXiv

BT - Lock in feedback in sequential experiments

PB - arXiv.org

ER -

Kaptein M, Iannuzzi D. Lock in feedback in sequential experiments. arXiv.org. 2015 Feb 2. (arXiv).