東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

FindBook

Google Book

Amazon

博客來

Online Resource Allocation and its Applications.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Online Resource Allocation and its Applications./
作者:	Zhu, Qiuyu.
面頁冊數:	1 online resource (121 pages)
附註:	Source: Dissertations Abstracts International, Volume: 84-04, Section: B.
Contained By:	Dissertations Abstracts International84-04B.
標題:	Dynamic programming. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=29352979click for full text (PQDT)
ISBN:	9798352683385

Online Resource Allocation and its Applications.
Zhu, Qiuyu.

Online Resource Allocation and its Applications. - 1 online resource (121 pages)

Source: Dissertations Abstracts International, Volume: 84-04, Section: B.

Thesis (Ph.D.)--National University of Singapore (Singapore), 2022.

Includes bibliographical references

Online resource allocation (ORA) is one of the most important problems in operations research. This thesis focus on algorithm design for different ORA models.In Chapter 2, we study the online resource allocation problem in which the resources are substitutable in two directions. To tackle the complicated substitution effect introduced by the multidimensional substitution, we proposed the Frontier Inventory Balancing (FIB) algorithm motivated by the dynamic programming formulation and a closedform solution of the linear programming approximation. We provide comprehensive competitive ratio analyses and extensive numerical studies for the proposed algorithms. Simulation studies show that our algorithm outperforms other state-of-art algorithms. In Chapter 3, we study the previous problem further by introducing 'learning' into the setting. Under the new setting, the arrival information is not known to the decision-maker. We generalize the FIB algorithm to the new setting and provide some theoretical results. The unknown arrival probability brings extra difficulties to the analyses. Extensive numerical studies are provided to compare the performance of different algorithms.The learning involved in Chapter 3 is limited because the decision-maker's action does not affect the learning. In this regard, we study another resource allocation model -multi-armed bandit (MAB)- in Chapter 4 and Chapter 5. MAB is a classical problem that exemplifies the exploration-exploitation trade-off. Standard formulations of MAB do not take into account risk. In online decision making systems, risk is a primary concern. In this regard, the mean-variance and CVaR risk measures are the most common objective functions. Existing algorithms for risk-aware MAB have unrealistic assumptions on the reward distributions. We develop Thompson Sampling-style algorithms for mean-variance and CVaR MAB, and provide comprehensive regret analyses. Our algorithms achieve the best known regret bounds for risk-aware MABs and also attain the information-theoretic bounds in some parameter regimes. Empirical simulations show that our algorithms significantly outperform existing LCB-based algorithms.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2023

Mode of access: World Wide Web

ISBN: 9798352683385Subjects--Topical Terms:

641303
Dynamic programming.
Index Terms--Genre/Form:

542853
Electronic books.

Online Resource Allocation and its Applications.
LDR:03509nmm a2200397K 4500 001 2356145
005 20230612071829.5
006 m o d
007 cr mn ---uuuuu
008 241011s2022 xx obm 000 0 eng d
020 $a 9798352683385
035 $a (MiAaPQ)AAI29352979
035 $a (MiAaPQ)USingapore216501
035 $a AAI29352979
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Zhu, Qiuyu. $3 3696618
245 1 0 $a Online Resource Allocation and its Applications.
264 0 $c 2022
300 $a 1 online resource (121 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 84-04, Section: B.
500 $a Advisor: Lim, Andrew.
502 $a Thesis (Ph.D.)--National University of Singapore (Singapore), 2022.
504 $a Includes bibliographical references
520 $a Online resource allocation (ORA) is one of the most important problems in operations research. This thesis focus on algorithm design for different ORA models.In Chapter 2, we study the online resource allocation problem in which the resources are substitutable in two directions. To tackle the complicated substitution effect introduced by the multidimensional substitution, we proposed the Frontier Inventory Balancing (FIB) algorithm motivated by the dynamic programming formulation and a closedform solution of the linear programming approximation. We provide comprehensive competitive ratio analyses and extensive numerical studies for the proposed algorithms. Simulation studies show that our algorithm outperforms other state-of-art algorithms. In Chapter 3, we study the previous problem further by introducing 'learning' into the setting. Under the new setting, the arrival information is not known to the decision-maker. We generalize the FIB algorithm to the new setting and provide some theoretical results. The unknown arrival probability brings extra difficulties to the analyses. Extensive numerical studies are provided to compare the performance of different algorithms.The learning involved in Chapter 3 is limited because the decision-maker's action does not affect the learning. In this regard, we study another resource allocation model -multi-armed bandit (MAB)- in Chapter 4 and Chapter 5. MAB is a classical problem that exemplifies the exploration-exploitation trade-off. Standard formulations of MAB do not take into account risk. In online decision making systems, risk is a primary concern. In this regard, the mean-variance and CVaR risk measures are the most common objective functions. Existing algorithms for risk-aware MAB have unrealistic assumptions on the reward distributions. We develop Thompson Sampling-style algorithms for mean-variance and CVaR MAB, and provide comprehensive regret analyses. Our algorithms achieve the best known regret bounds for risk-aware MABs and also attain the information-theoretic bounds in some parameter regimes. Empirical simulations show that our algorithms significantly outperform existing LCB-based algorithms.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2023
538 $a Mode of access: World Wide Web
650 4 $a Dynamic programming. $3 641303
650 4 $a Flexibility. $3 3560705
650 4 $a Design. $3 518875
650 4 $a Numerical analysis. $3 517751
650 4 $a Linear programming. $3 560448
650 4 $a Stochastic models. $3 764002
650 4 $a Drones. $3 3549538
650 4 $a Algorithms. $3 536374
650 4 $a Internet resources. $3 3696619
650 4 $a Car pools. $3 3683472
650 4 $a Competition. $3 537031
650 4 $a Applied mathematics. $3 2122814
650 4 $a Computer science. $3 523869
650 4 $a Mass communications. $3 3422380
650 4 $a Mathematics. $3 515831
650 4 $a Transportation. $3 555912
655 7 $a Electronic books. $2 lcsh $3 542853
690 $a 0389
690 $a 0364
690 $a 0984
690 $a 0338
690 $a 0708
690 $a 0405
690 $a 0709
710 2 $a ProQuest Information and Learning Co. $3 783688
710 2 $a National University of Singapore (Singapore). $3 3352228
773 0 $t Dissertations Abstracts International $g 84-04B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=29352979 $z click for full text (PQDT)