Cost-Sensitive Portfolio Selection via Deep Reinforcement Learning

Portfolio Selection is an important real-world financial task and has attracted extensive attention in artificial intelligence communities. This task, however, has two main difficulties: (i) the non-stationary price series and complex asset correlations make the learning of feature representation ve...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	IEEE transactions on knowledge and data engineering Ročník 34; číslo 1; s. 236 - 248
Hlavní autoři:	Zhang, Yifan, Zhao, Peilin, Wu, Qingyao, Li, Bin, Huang, Junzhou, Tan, Mingkui
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	New York IEEE 01.01.2022 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:	Artificial intelligence Correlation Cost analysis Costs Decision making Deep learning Feature extraction Investment Learning (artificial intelligence) Optimization Portfolio selection Portfolios Profitability reinforcement learning Representations Risk management Task analysis transaction cost
ISSN:	1041-4347, 1558-2191
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Abstract	Portfolio Selection is an important real-world financial task and has attracted extensive attention in artificial intelligence communities. This task, however, has two main difficulties: (i) the non-stationary price series and complex asset correlations make the learning of feature representation very hard; (ii) the practicality principle in financial markets requires controlling both transaction and risk costs. Most existing methods adopt handcraft features and/or consider no constraints for the costs, which may make them perform unsatisfactorily and fail to control both costs in practice. In this paper, we propose a cost-sensitive portfolio selection method with deep reinforcement learning. Specifically, a novel two-stream portfolio policy network is devised to extract both price series patterns and asset correlations, while a new cost-sensitive reward function is developed to maximize the accumulated return and constrain both costs via reinforcement learning. We theoretically analyze the near-optimality of the proposed reward, which shows that the growth rate of the policy regarding this reward function can approach the theoretical optimum. We also empirically evaluate the proposed method on real-world datasets. Promising results demonstrate the effectiveness and superiority of the proposed method in terms of profitability, cost-sensitivity and representation abilities.
AbstractList	Portfolio Selection is an important real-world financial task and has attracted extensive attention in artificial intelligence communities. This task, however, has two main difficulties: (i) the non-stationary price series and complex asset correlations make the learning of feature representation very hard; (ii) the practicality principle in financial markets requires controlling both transaction and risk costs. Most existing methods adopt handcraft features and/or consider no constraints for the costs, which may make them perform unsatisfactorily and fail to control both costs in practice. In this paper, we propose a cost-sensitive portfolio selection method with deep reinforcement learning. Specifically, a novel two-stream portfolio policy network is devised to extract both price series patterns and asset correlations, while a new cost-sensitive reward function is developed to maximize the accumulated return and constrain both costs via reinforcement learning. We theoretically analyze the near-optimality of the proposed reward, which shows that the growth rate of the policy regarding this reward function can approach the theoretical optimum. We also empirically evaluate the proposed method on real-world datasets. Promising results demonstrate the effectiveness and superiority of the proposed method in terms of profitability, cost-sensitivity and representation abilities.
Author	Zhao, Peilin Huang, Junzhou Tan, Mingkui Wu, Qingyao Li, Bin Zhang, Yifan
Author_xml	– sequence: 1 givenname: Yifan orcidid: 0000-0002-2125-1074 surname: Zhang fullname: Zhang, Yifan email: sezyifan@mail.scut.edu.cn organization: South China University of Technology, Guangzhou, China – sequence: 2 givenname: Peilin orcidid: 0000-0001-8543-3953 surname: Zhao fullname: Zhao, Peilin email: peilinzhao@hotmail.com organization: AI Lab, Tencent, Shenzhen, China – sequence: 3 givenname: Qingyao orcidid: 0000-0002-6771-3932 surname: Wu fullname: Wu, Qingyao email: qyw@scut.edu.cn organization: South China University of Technology, Guangzhou, China – sequence: 4 givenname: Bin surname: Li fullname: Li, Bin email: binli.whu@whu.edu.cn organization: Wuhan University, Wuhan, China – sequence: 5 givenname: Junzhou surname: Huang fullname: Huang, Junzhou email: joehhuang@tencent.com organization: AI Lab, Tencent, Shenzhen, China – sequence: 6 givenname: Mingkui orcidid: 0000-0001-8856-756X surname: Tan fullname: Tan, Mingkui email: mingkuitan@scut.edu.cn organization: South China University of Technology, Guangzhou, China
BookMark	eNp9kE1PAjEQhhuDiYD-AONlE8-LM9uybY8K-BFJNILnTVlmTcnSYltI_PdCMB48eJo5vM-8mafHOs47YuwSYYAI-mb-PJ4MCihgUGipJcAJ6-JwqPICNXb2OwjMBRfyjPViXAGAkgq77G7kY8pn5KJNdkfZqw-p8a312YxaqpP1LttZk42JNtkbWdf4UNOaXMqmZIKz7uOcnTamjXTxM_vs_X4yHz3m05eHp9HtNK8LzVOuBC5IKg0LYxRyU6tiaQpsOKKRS6OphFLrBdQLWipDig953YhSIDUkQRDvs-vj3U3wn1uKqVr5bXD7yqooQSop5VDsU3hM1cHHGKipNsGuTfiqEKqDquqgqjqoqn5U7Rn5h6ltMoffUzC2_Ze8OpKWiH6bNHAUqPg3ytZ4TA
CODEN	ITKEEH
CitedBy_id	crossref_primary_10_1049_gtd2_12040 crossref_primary_10_1080_24725854_2021_1973156 crossref_primary_10_1109_ACCESS_2022_3166599 crossref_primary_10_1007_s10614_024_10786_z crossref_primary_10_1109_ACCESS_2021_3068768 crossref_primary_10_1109_LSP_2023_3296913 crossref_primary_10_1007_s10489_024_06107_4 crossref_primary_10_1007_s00766_025_00436_7 crossref_primary_10_1186_s13059_022_02692_0 crossref_primary_10_1007_s10462_024_11066_w crossref_primary_10_1109_TKDE_2025_3545115 crossref_primary_10_1145_3733714 crossref_primary_10_4018_IJITWE_332777 crossref_primary_10_1109_TKDE_2023_3251721 crossref_primary_10_1007_s10489_025_06242_6 crossref_primary_10_1016_j_orl_2023_10_011 crossref_primary_10_1007_s11067_024_09663_x crossref_primary_10_1109_TNNLS_2022_3231917 crossref_primary_10_1016_j_knosys_2022_109650 crossref_primary_10_1109_TKDE_2023_3250472 crossref_primary_10_1016_j_neunet_2021_09_002 crossref_primary_10_1016_j_eswa_2024_124193 crossref_primary_10_1109_TKDE_2023_3301036 crossref_primary_10_1016_j_eswa_2023_120297 crossref_primary_10_1016_j_ins_2022_08_074
Cites_doi	10.1109/ICCV.2019.00719 10.1016/B978-1-55860-335-6.50021-0 10.1007/978-3-540-77226-2_102 10.1109/CVPR.2015.7298965 10.1023/A:1010884214864 10.1609/aaai.v27i1.8693 10.2307/2975974 10.1007/978-3-030-32239-7_40 10.1142/9789814293501_0015 10.1007/s10479-016-2316-z 10.1007/978-3-540-87987-9_13 10.1609/aaai.v28i1.8907 10.1109/TKDE.2016.2563433 10.1111/1467-9965.00058 10.1613/jair.1336 10.1109/TKDE.2018.2826011 10.1145/2512962 10.1109/IHMSC.2013.84 10.1109/CVPR42600.2020.00545 10.1080/14697688.2017.1357831 10.1609/aaai.v29i1.9453 10.1145/2229012.2229022 10.32657/10356/90191 10.1109/CVPR42600.2020.01264 10.1109/TNNLS.2016.2522401 10.1145/3219819.3220113 10.1016/j.dss.2004.08.010 10.1007/978-3-662-54970-4_28 10.1145/1143844.1143846 10.1109/TIP.2020.3006377 10.1109/CVPR42600.2020.01030 10.1214/aop/1176991793 10.1109/72.935097 10.1287/mnsc.2013.1838 10.1109/TKDE.2019.2955078 10.1145/2435209.2435213 10.1016/j.ijforecast.2003.09.015 10.1038/nature14236 10.1145/3219819.3219948 10.1002/j.1538-7305.1956.tb03809.x 10.1145/3274643 10.1080/14697688.2011.570368 10.1145/2939672.2939861 10.1007/s10994-012-5281-z 10.1145/3269206.3269290 10.1016/0196-8858(86)90029-1 10.1609/aaai.v31i1.10728 10.1162/neco.1997.9.8.1735 10.1038/Nature14539 10.1609/aaai.v33i01.33011134 10.1524/stnd.2007.25.1.63 10.1007/978-3-642-36288-0_17 10.1002/(SICI)1099-131X(1998090)17:5/6<441::AID-FOR707>3.0.CO;2-# 10.1109/CVPR.2017.113 10.1007/978-1-4615-5625-1_10
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022
DBID	97E RIA RIE AAYXX CITATION 7SC 7SP 8FD JQ2 L7M L~C L~D
DOI	10.1109/TKDE.2020.2979700
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE/IET Electronic Library CrossRef Computer and Information Systems Abstracts Electronics & Communications Abstracts Technology Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional
DatabaseTitle	CrossRef Technology Research Database Computer and Information Systems Abstracts – Academic Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Professional
DatabaseTitleList	Technology Research Database
Database_xml	– sequence: 1 dbid: RIE name: IEEE/IET Electronic Library (IEL) (UW System Shared) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering Computer Science
EISSN	1558-2191
EndPage	248
ExternalDocumentID	10_1109_TKDE_2020_2979700 9031418
Genre	orig-research
GrantInformation_xml	– fundername: Guangdong Provincial Scientific and Technological Funds grantid: 2018B010107001; 2018B010108002 – fundername: Program for Guangdong Introducing Innovative and Enterpreneurial Teams grantid: 2017ZT07X183 – fundername: National Natural Science Foundation of China grantid: 61876208; 71971164 funderid: 10.13039/501100001809 – fundername: Fundamental Research Funds for the Central Universities grantid: D2191240 funderid: 10.13039/501100012226 – fundername: NSFC grantid: 61836003 – fundername: Tencent AI Lab Rhino-Bird Focused Research Program grantid: JR201902 – fundername: Pearl River S and T Nova Program of Guangzhou grantid: 201806010081 funderid: 10.13039/501100009334
GroupedDBID	-~X .DC 0R~ 29I 4.4 5GY 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACGFO ACIWK AENEX AGQYO AHBIQ AKJIK AKQYR ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD F5P HZ~ IEDLZ IFIPE IPLJI JAVBF LAI M43 MS~ O9- OCL P2P PQQKQ RIA RIE RNS RXW TAE TN5 UHB 1OL 5VS 9M8 AAYXX ABFSI AETIX AGSQL AI. AIBXA ALLEH CITATION E.L H~9 ICLAB IFJZH RNI RZB TAF VH1 7SC 7SP 8FD JQ2 L7M L~C L~D
ID	FETCH-LOGICAL-c293t-841be7890baa813ac82da21f311a7da9e60699b0cbed8ae8353cf4641efe704e3
IEDL.DBID	RIE
ISICitedReferencesCount	79
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000728576400016&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	1041-4347
IngestDate	Sun Jun 29 14:10:08 EDT 2025 Sat Nov 29 02:36:02 EST 2025 Tue Nov 18 22:35:39 EST 2025 Wed Aug 27 05:11:53 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	1
Language	English
License	https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c293t-841be7890baa813ac82da21f311a7da9e60699b0cbed8ae8353cf4641efe704e3
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ORCID	0000-0002-2125-1074 0000-0001-8543-3953 0000-0001-8856-756X 0000-0002-6771-3932
PQID	2607877754
PQPubID	85438
PageCount	13
ParticipantIDs	ieee_primary_9031418 crossref_primary_10_1109_TKDE_2020_2979700 proquest_journals_2607877754 crossref_citationtrail_10_1109_TKDE_2020_2979700
PublicationCentury	2000
PublicationDate	2022-01-01
PublicationDateYYYYMMDD	2022-01-01
PublicationDate_xml	– month: 01 year: 2022 text: 2022-01-01 day: 01
PublicationDecade	2020
PublicationPlace	New York
PublicationPlace_xml	– name: New York
PublicationTitle	IEEE transactions on knowledge and data engineering
PublicationTitleAbbrev	TKDE
PublicationYear	2022
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref13 ref57 ref12 ref56 ref15 ref14 ref58 ref53 Sutskever (ref60) ref52 Van Den Oord (ref63) ref11 ref10 ref54 ref17 ref16 ref19 Yu (ref66) ref18 Schulman (ref55) ref46 Huang (ref30) ref45 ref48 Cao (ref8) ref47 Guo (ref22) 2018 ref42 ref41 ref44 ref43 Neuneier (ref49) ref9 Li (ref37) ref4 ref3 Sikdar (ref59) ref6 Vajda (ref62) ref40 Jiang (ref31) 2017 Bai (ref5) 2018 ref35 ref34 ref36 ref75 ref74 ref33 ref32 ref2 Neuneier (ref51) Sutton (ref61) ref39 ref38 ref71 ref70 ref73 ref72 Bradbury (ref7) 2016 ref24 Wang (ref64) ref68 ref23 ref67 ref26 ref25 ref69 ref21 ref65 ref28 Neuneier (ref50) ref27 Abadi (ref1) ref29 Guo (ref20)
References_xml	– ident: ref68 doi: 10.1109/ICCV.2019.00719 – ident: ref24 doi: 10.1016/B978-1-55860-335-6.50021-0 – start-page: 952 volume-title: Proc. 8th Int. Conf. Neural Inf. Process. Syst. ident: ref50 article-title: Optimal asset allocation using adaptive dynamic programming – start-page: 1057 volume-title: Proc. 12th Int. Conf. Neural Inf. Process. Syst. ident: ref61 article-title: Policy gradient methods for reinforcement learning with function approximation – ident: ref40 doi: 10.1007/978-3-540-77226-2_102 – ident: ref42 doi: 10.1109/CVPR.2015.7298965 – start-page: 936 volume-title: Proc. Int. Conf. Neural Inf. Process. Syst. ident: ref51 article-title: Enhancing Q-learning for optimal asset allocation – ident: ref19 doi: 10.1023/A:1010884214864 – ident: ref13 doi: 10.1609/aaai.v27i1.8693 – ident: ref43 doi: 10.2307/2975974 – ident: ref74 doi: 10.1007/978-3-030-32239-7_40 – ident: ref12 doi: 10.1142/9789814293501_0015 – ident: ref75 doi: 10.1007/s10479-016-2316-z – ident: ref23 doi: 10.1007/978-3-540-87987-9_13 – ident: ref14 doi: 10.1609/aaai.v28i1.8907 – ident: ref29 doi: 10.1109/TKDE.2016.2563433 – ident: ref26 doi: 10.1111/1467-9965.00058 – ident: ref6 doi: 10.1613/jair.1336 – start-page: 735 volume-title: Proc. Int. Conf. Neural Inf. Process. Syst. ident: ref20 article-title: NAT: Neural architecture transformer for accurate and compact architectures – ident: ref70 doi: 10.1109/TKDE.2018.2826011 – year: 2018 ident: ref5 article-title: An empirical evaluation of generic convolutional and recurrent networks for sequence modeling – ident: ref38 doi: 10.1145/2512962 – start-page: 2165 volume-title: Proc. AAAI Conf. Artif. Intell. ident: ref59 article-title: Top-trading-cycles mechanisms with acceptable bundles – start-page: 273 volume-title: Proc. Int. Conf. Mach. Learn. ident: ref37 article-title: On-line portfolio selection with moving average reversion – ident: ref18 doi: 10.1109/IHMSC.2013.84 – ident: ref21 doi: 10.1109/CVPR42600.2020.00545 – ident: ref39 doi: 10.1080/14697688.2017.1357831 – start-page: 1031 volume-title: Proc. Int. Conf. Neural Inf. Process. Syst. ident: ref49 article-title: Risk sensitive reinforcement learning – ident: ref56 doi: 10.1609/aaai.v29i1.9453 – start-page: 265 volume-title: Proc. USENIX Symp. Operating Syst. Des. Implementation ident: ref1 article-title: TensorFlow: A system for large-scale machine learning – ident: ref4 doi: 10.1145/2229012.2229022 – ident: ref41 doi: 10.32657/10356/90191 – ident: ref10 doi: 10.1109/CVPR42600.2020.01264 – ident: ref15 doi: 10.1109/TNNLS.2016.2522401 – ident: ref16 doi: 10.1145/3219819.3220113 – ident: ref17 doi: 10.1016/j.dss.2004.08.010 – ident: ref54 doi: 10.1007/978-3-662-54970-4_28 – ident: ref2 doi: 10.1145/1143844.1143846 – ident: ref73 doi: 10.1109/TIP.2020.3006377 – ident: ref67 doi: 10.1109/CVPR42600.2020.01030 – ident: ref3 doi: 10.1214/aop/1176991793 – ident: ref45 doi: 10.1109/72.935097 – ident: ref48 doi: 10.1287/mnsc.2013.1838 – ident: ref72 doi: 10.1109/TKDE.2019.2955078 – volume-title: Proc. Int. Conf. Mach. Learn. ident: ref55 article-title: Trust region policy optimization – start-page: 178 volume-title: Proc. Int. Joint Conf. Artif. Intell. ident: ref30 article-title: Semi-universal portfolios with transaction costs – ident: ref35 doi: 10.1145/2435209.2435213 – ident: ref28 doi: 10.1016/j.ijforecast.2003.09.015 – ident: ref44 doi: 10.1038/nature14236 – year: 2017 ident: ref31 article-title: A deep reinforcement learning framework for the financial portfolio management problem publication-title: arXiv:1706.10059 – start-page: 3104 volume-title: Proc. 27th Int. Conf. Neural Inf. Process. Syst. ident: ref60 article-title: Sequence to sequence learning with neural networks – ident: ref71 doi: 10.1145/3219819.3219948 – year: 2018 ident: ref22 article-title: Robust log-optimal strategy with reinforcement learning – ident: ref32 doi: 10.1002/j.1538-7305.1956.tb03809.x – ident: ref25 doi: 10.1145/3274643 – ident: ref52 doi: 10.1080/14697688.2011.570368 – start-page: 125 volume-title: Proc. ISCA Speech Synthesis Workshop ident: ref63 article-title: WaveNet: A generative model for raw audio – ident: ref69 doi: 10.1145/2939672.2939861 – start-page: 1995 volume-title: Proc. 33rd Int. Conf. Mach. Learn. ident: ref64 article-title: Dueling network architectures for deep reinforcement learning – ident: ref36 doi: 10.1007/s10994-012-5281-z – start-page: 719 volume-title: Proc. Prague Stochastics ident: ref62 article-title: Analysis of semi-log-optimal investment strategies – ident: ref65 doi: 10.1145/3269206.3269290 – ident: ref11 doi: 10.1016/0196-8858(86)90029-1 – ident: ref57 doi: 10.1609/aaai.v31i1.10728 – ident: ref27 doi: 10.1162/neco.1997.9.8.1735 – ident: ref34 doi: 10.1038/Nature14539 – ident: ref58 doi: 10.1609/aaai.v33i01.33011134 – ident: ref53 doi: 10.1524/stnd.2007.25.1.63 – start-page: 1774 volume-title: Proc. 33rd Int. Conf. Neural Inf. Process. Syst. ident: ref8 article-title: Multi-marginal Wasserstein GAN – ident: ref9 doi: 10.1007/978-3-642-36288-0_17 – ident: ref46 doi: 10.1002/(SICI)1099-131X(1998090)17:5/6<441::AID-FOR707>3.0.CO;2-# – volume-title: Proc. Int. Conf. Learn. Representations ident: ref66 article-title: Multi-scale context aggregation by dilated convolutions – year: 2016 ident: ref7 article-title: Quasi-recurrent neural networks – ident: ref33 doi: 10.1109/CVPR.2017.113 – ident: ref47 doi: 10.1007/978-1-4615-5625-1_10
SSID	ssj0008781
Score	2.6347182
Snippet	Portfolio Selection is an important real-world financial task and has attracted extensive attention in artificial intelligence communities. This task, however,...
SourceID	proquest crossref ieee
SourceType	Aggregation Database Enrichment Source Index Database Publisher
StartPage	236
SubjectTerms	Artificial intelligence Correlation Cost analysis Costs Decision making Deep learning Feature extraction Investment Learning (artificial intelligence) Optimization Portfolio selection Portfolios Profitability reinforcement learning Representations Risk management Task analysis transaction cost
Title	Cost-Sensitive Portfolio Selection via Deep Reinforcement Learning
URI	https://ieeexplore.ieee.org/document/9031418 https://www.proquest.com/docview/2607877754
Volume	34
WOSCitedRecordID	wos000728576400016&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE/IET Electronic Library (IEL) (UW System Shared) customDbUrl: eissn: 1558-2191 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0008781 issn: 1041-4347 databaseCode: RIE dateStart: 19890101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LSwMxEB5q8aAHH1WxWiUHT-K2m2y2SY5aFUEo4gO8LdnsrBRKt_T1-03StBQUwdsekmWZ2cwrM98HcMVoYhxSVsQRVcRtDBAp1k2jHNPEuhydSrMkmxD9vvz8VC81uFnPwiCibz7Dtnv0d_lFZeauVNZRDmudyi3YEkIsZ7XWVlcKT0hqswubEyVchBtMGqvO-_P9g80EWdxmSijhhtk2fJAnVflhib17edz_34cdwF4II8ntUu-HUMNRA_ZXFA0knNgG7G7gDR7BXa-azqI317PurBxxbaRlNRxU5M3T4VgdkcVAk3vEMXlFD6pqfP2QBBzWr2P4eHx47z1FgUQhMtaTzyLJaY5u2jXXWtJEG8kKzWiZUKpFoRXaDEapPDY5FlKjDcgSU_Iup1iiiDkmJ1AfVSM8BaLTEpnRvGCJjbty97pY8sLaWixFEXebEK_EmpmAMO6ILoaZzzRilTlNZE4TWdBEE67XW8ZLeI2_Fh850a8XBqk3obXSXRYO4DSzaZpwUIcpP_t91znsMDfJ4KspLajPJnO8gG2zmA2mk0v_b30Dw_rKWg
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LSwMxEB58gXrwLdZnDp7E1U022yRHrYpiLWIreFuy2VkpSFfs4_ebpGkpKIK3PSTLMrOZV2a-D-CU0cQ4pKyII6qI2xggUqyeRjmmiXU5OpVmTDYhWi359qae5-B8OguDiL75DC_co7_LLyozdKWyS-Ww1qmch8WUc0bH01pTuyuFpyS1-YXNihIuwh0mjdVl5_Hm1uaCLL5gSijhxtlmvJCnVflhi72DuVv_36dtwFoIJMnVWPObMIe9LVifkDSQcGa3YHUGcXAbrhtVfxC1Xde6s3PENZKW1Ue3Im1PiGO1REZdTW4QP8kLelhV4yuIJCCxvu_A691tp3EfBRqFyFhfPogkpzm6eddca0kTbSQrNKNlQqkWhVZocxil8tjkWEiNNiRLTMnrnGKJIuaY7MJCr-rhHhCdlsiM5gVLbOSVu9fFkhfW2mIpirheg3gi1swEjHFHdfGR-VwjVpnTROY0kQVN1OBsuuVzDLDx1-JtJ_rpwiD1GhxOdJeFI9jPbKImHNhhyvd_33UCy_edp2bWfGg9HsAKc3MNvrZyCAuDryEewZIZDbr9r2P_n30DdNPNoQ
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Cost-Sensitive+Portfolio+Selection+via+Deep+Reinforcement+Learning&rft.jtitle=IEEE+transactions+on+knowledge+and+data+engineering&rft.au=Zhang%2C+Yifan&rft.au=Zhao%2C+Peilin&rft.au=Wu%2C+Qingyao&rft.au=Li%2C+Bin&rft.date=2022-01-01&rft.pub=The+Institute+of+Electrical+and+Electronics+Engineers%2C+Inc.+%28IEEE%29&rft.issn=1041-4347&rft.eissn=1558-2191&rft.volume=34&rft.issue=1&rft.spage=236&rft_id=info:doi/10.1109%2FTKDE.2020.2979700&rft.externalDBID=NO_FULL_TEXT
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1041-4347&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1041-4347&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1041-4347&client=summon