Scalarized lower upper confidence bound algorithm

M.M. Drugan

doi:10.1007/978-3-319-19084-6_21

Scalarized lower upper confidence bound algorithm

M.M. Drugan

Security

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

Abstract

Multi-objective evolutionary optimisation algorithms and stochastic multi-armed bandits techniques are combined in designing stochastic multi-objective multi-armed bandits (MOMAB) with an efficient exploration and exploitation trade-off. Lower upper confidence bound (LUCB) focuses on sampling the arms that are most probable to be misclassified (i.e., optimal or suboptimal arms) in order to identify the set of best arms aka the Pareto front. Our scalarized multi-objective LUCB (sMO-LUCB) is an adaptation of LUCB to reward vectors. Preliminary empirical results show good performance of the proposed algorithm on a bi-objective environment.

Original language	English
Title of host publication	Learning and Intelligent Optimization
Subtitle of host publication	9th International Conference, LION 9, Lille, France, January 12-15, 2015. Revised Selected Papers
Editors	Cl. Dhaenens, L. Jourdan, M.-E. Marmion
Place of Publication	Berlin
Publisher	Springer
Pages	229-235
ISBN (Print)	978-3-319-19083-9
DOIs	https://doi.org/10.1007/978-3-319-19084-6_21
Publication status	Published - 2015

Publication series

Name	Lecture Notes in Computer Science
Volume	8994

Access to Document

10.1007/978-3-319-19084-6_21

Cite this

Drugan, M. M. (2015). Scalarized lower upper confidence bound algorithm. In C. Dhaenens, L. Jourdan, & M.-E. Marmion (Eds.), Learning and Intelligent Optimization : 9th International Conference, LION 9, Lille, France, January 12-15, 2015. Revised Selected Papers (pp. 229-235). (Lecture Notes in Computer Science; Vol. 8994). Springer. https://doi.org/10.1007/978-3-319-19084-6_21

@inproceedings{99d886d7196a4cdcbf6da00a55cb2657,

title = "Scalarized lower upper confidence bound algorithm",

abstract = "Multi-objective evolutionary optimisation algorithms and stochastic multi-armed bandits techniques are combined in designing stochastic multi-objective multi-armed bandits (MOMAB) with an efficient exploration and exploitation trade-off. Lower upper confidence bound (LUCB) focuses on sampling the arms that are most probable to be misclassified (i.e., optimal or suboptimal arms) in order to identify the set of best arms aka the Pareto front. Our scalarized multi-objective LUCB (sMO-LUCB) is an adaptation of LUCB to reward vectors. Preliminary empirical results show good performance of the proposed algorithm on a bi-objective environment.",

author = "M.M. Drugan",

year = "2015",

doi = "10.1007/978-3-319-19084-6_21",

language = "English",

isbn = "978-3-319-19083-9",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "229--235",

editor = "Cl. Dhaenens and L. Jourdan and M.-E. Marmion",

booktitle = "Learning and Intelligent Optimization",

address = "Germany",

}

Drugan, MM 2015, Scalarized lower upper confidence bound algorithm. in C Dhaenens, L Jourdan & M-E Marmion (eds), Learning and Intelligent Optimization : 9th International Conference, LION 9, Lille, France, January 12-15, 2015. Revised Selected Papers. Lecture Notes in Computer Science, vol. 8994, Springer, Berlin, pp. 229-235. https://doi.org/10.1007/978-3-319-19084-6_21

Scalarized lower upper confidence bound algorithm. / Drugan, M.M.
Learning and Intelligent Optimization : 9th International Conference, LION 9, Lille, France, January 12-15, 2015. Revised Selected Papers. ed. / Cl. Dhaenens; L. Jourdan; M.-E. Marmion. Berlin: Springer, 2015. p. 229-235 (Lecture Notes in Computer Science; Vol. 8994).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Scalarized lower upper confidence bound algorithm

AU - Drugan, M.M.

PY - 2015

Y1 - 2015

N2 - Multi-objective evolutionary optimisation algorithms and stochastic multi-armed bandits techniques are combined in designing stochastic multi-objective multi-armed bandits (MOMAB) with an efficient exploration and exploitation trade-off. Lower upper confidence bound (LUCB) focuses on sampling the arms that are most probable to be misclassified (i.e., optimal or suboptimal arms) in order to identify the set of best arms aka the Pareto front. Our scalarized multi-objective LUCB (sMO-LUCB) is an adaptation of LUCB to reward vectors. Preliminary empirical results show good performance of the proposed algorithm on a bi-objective environment.

AB - Multi-objective evolutionary optimisation algorithms and stochastic multi-armed bandits techniques are combined in designing stochastic multi-objective multi-armed bandits (MOMAB) with an efficient exploration and exploitation trade-off. Lower upper confidence bound (LUCB) focuses on sampling the arms that are most probable to be misclassified (i.e., optimal or suboptimal arms) in order to identify the set of best arms aka the Pareto front. Our scalarized multi-objective LUCB (sMO-LUCB) is an adaptation of LUCB to reward vectors. Preliminary empirical results show good performance of the proposed algorithm on a bi-objective environment.

U2 - 10.1007/978-3-319-19084-6_21

DO - 10.1007/978-3-319-19084-6_21

M3 - Conference contribution

SN - 978-3-319-19083-9

T3 - Lecture Notes in Computer Science

SP - 229

EP - 235

BT - Learning and Intelligent Optimization

A2 - Dhaenens, Cl.

A2 - Jourdan, L.

A2 - Marmion, M.-E.

PB - Springer

CY - Berlin

ER -

Scalarized lower upper confidence bound algorithm

Abstract

Publication series

Access to Document

Fingerprint

Cite this