CC BY
3Регион и мир, 2019, № 4
Uhfc tfhptUmjm^mU niumgnLdp U qpmUg
^^pmnnLpjnLUUtp^ ^tUrnpnUm^mU pmU^tpnL^
Umpqujmh U. Z-
Zm]mummh]i ^Uhmmljmh ShmhumqJimmljmh Ztuiim[umpmh (Ztujmummh, bplimh)
and.sargsyan@yahoo.com
4^nnpn2 pmnhp Jhb m^jmihhp, mphhumm^mh pmhmlmhntpjnth, tfhphhmjmlmh nLuntgntJ, Ihhmpnhmlmh pmh^hp
Большие данные, машинное обучение и их применение в центральных банках
Саргсян А. О.
Армянский государственный экономический университет (Армения, Ереван)
and.sargsyan@yahoo.com
Резюме: Развитие компьютерных технологий и снижение стоимости хранения данных привело к сбору и сохранению массовых электронных данных организациями. Трудно, а иногда и невозможно, проанализировать большие данные с помощью существующих традиционных методов. Для их анализа созданы новые статистические модели, которые строятся на основе данных и делают прогнозы. В статье описываются большие данные, машинное обучение, их применение и проблемы связанные с внедрением в центральных банках. Ключевые слова: центральные банки, большие данные, искусственный интеллект, машинное обучение
Big Data, Machine Learning and their applications in Central Banks
Sargsyan A. H.
Armenian State University of Economics (Armenia, Yerevan)
and.sargsyan@yahoo.com
Abstract: The development of computer technology and the reduction of data storage cost resulted in the collection and maintenance of mass electronic data by organizations. It is difficult, sometimes impossible, to analyze big data through existing traditional methods. New statistical models have been developed which are built from the given data and make predictions about the future. The article describes big data and machine learning, their applications and challenges associated with the implementation in Central Banks. Keywords: central banks, big data, artificial intelligence, machine learning
ZmJm^mpq^mj^h mh^h^lmj^ qmpqmgntJp U m^jmihhp^ ^mh^mhtfmh qh^ ^hgnttfp hmhqhgp^h Imqtfmlhp^nLpjnthhhp^ ^nqtf^g hulmjm&m^mi_ tl_hlrnpnhmj^h m^mihhp^ hm-^mpmqptfmh U ^mh^mhtfmh: Ujq Ьш^шф m^jm^htpp qnjntpjnth nthhgnq m^mhqm^mh tfhpnqhhpn^ ^hpjnL&hip pmpq t, hpphJh t]_ mhhhmp: Ipmhg ^hpint&tfmh hmtfmp umhqb-hh ^^^m^mqpm^mh hnp tfnqhihhp, npnhp Imnntg^ntJ hh mp^mb m^mihhp^ h^tfmh ^pm U Imh^mmhuntJhhp ^mrnmpntJ mqmqmj^ ^hpmphpjmp Znq^m&nttf hhplmjmg^nttf hh Jhb m^mihhp^, tfhphhmjmlmh ntuntgtfmh umhtfm-hnttfhhpp, qpmhg l^pmnntpjnthhhpp U hhpqp-Jmh hhm Цшц^шЬ tfmpmmhpm^hphhpp Ihhrnpnhmlmh pmh^hpntJ:
Zm2^nqm^mh mh^h^lmj^ U ^hmhphhm^ qmpqmgtfmhp qntqphpmg rnmpphp mqpjntphh-p^g qhhhpmg^nq U ummg^nq mhqhlmm^ntp-jmh Ьш^шщ hplpш¿шфшlшh ^pnqphu^mjn^
m^hg: mnmgmgphg mhhpmdh2mnLpjnLh umhqfchpi. htfmh m^mihhp^ ^mh^mhtfmh U J^ltfmh hnp Jh^mh^qtfhhp: Ujq Jh^mh^qtf-hhp^ hhpqpnttfp mjuop qmh^niJ t fcn2np ^mq-Jmlhp^nLpjnthhhp^, mjq p-^niJ hmU 4hhmpn-hm^mh pmh^hp^ ni^mqpntpjmh IhhrnpnhntJ: 4hhrnpnhmlmh pmh^hpp, oqmmqnp&hjn^ Jh-ЬшЬш^ш^ U rnmpphp mqpjntphhp^g ummg^nq m^jmihhpp, ^npA-niJ hh Immmph^ ш^Ьф
Imh^mrnhuntJhhp U tf2mlhl_ mpqjnthm^hm qpmtfm^mplmj^h pmqmpm^m-hntpjnth:
Uhb hh hmJmp^ntJ tfh&m&m^mp pmpq U mpmq m^nq m^mihhpp, npnhp ^mhmhgntJ hh mnmgmqhJ mh^h^lm U mh^hnjnq^m ^hфnp-Jmg^mh hm^mpmqphjnL, ^mh^mhhjnt, pm2-^h^nt, Imnm^mphpi. U ^hpinLfchpi. hmJmp1:
1 St u Demystifying Big Data: A Practical Guide To Transforming The Business of Government, 2012
Uhb s^jmLühp (Big data) shptf^üp mnmg^ü müqmJ l^pmnhj t O'Reilly Media hpmsmpml-^nrpjmü tfmpphp^üq^ pmdü^ süophü ftnghp Umqnqmup 2005 p^mlmü^ü: Im ^hpmphpnrtf tp m^jm^ühp^ mjüi^u^ Jhb pmqdmpjmüp, npp qphph mühümp tp Imnm^mphj U U^mlhL oqsmqnpbhLn^ p^qühu ^hpLnrbnLpjmü qnpb^p-ühpp2: Qümjmb üpmü, np Jhb m^jm^ühp hmu-Imgnrpjnrüp uluhj t 2p2mümn^hj qhnUu 2005 p^mlmü^g, qpm ülmsJmJp hhsmppppnrp-jniüp mnm^hj Jhb shtfin^ uluhj t m^hj 2012 p^m^mü^g: 100 90 30 70 Ë0 50 40
OOOOOOT-T-T-T-T-T-*-T-T-
ooooooooooooooo
r^íNr^íMnr^íMnr^íMnr^íMCNrj
Q-àmiqminljhp 1. "Bigdata"inhpiijiüji npnüifmüpphüqp 2004-2018pp. pus Google Trends oü[mjü qnpbjipji3
Uhb s^jmiühpp un^npmpmp pünp^^nrtf hü hphp mnmügpühpn^ (3V Jnqh^)4 •
• bm^mj (volume)
• mpmqntpjntü (velocity)
• pmqJmqmüntpjntü (variety) Üm^mjp ü2mümlntJ t, np mjq s^jmLühpp
^hpmphpnrJ hü hulmjmbm^mj ^ü^nptfmg^mj^ ^hpLntbntpjmüp un^npmpmp ulumb 10 U m^h-L^ shpmpmjp^g: IBM-^ hhmmqnmnLpjntüühpp gnrjg hü sm)^u, np m2^mph^ pnLnp s^jmL-ühp^ 90%-p ushqb^nrtf hü üm^npqnq 2 sm-p^ühp^ püpmgpnrtf5: Uju^üpü nrühüp s^jmL-ühp^ bm^m^ hplpm^m^mlmü ipnqphu^m-jn^ m£, U m^^ shlip tf^mjü m^h^ t mpm-qmüntJ:
2 Shu Gali Halevi, Henk Moed. 2012. The evolution of big data as a research and scientific topic: Overview of the literature
3 ^bm^mmltpp Imnnig^hl. t hhq^üml^ InqJ^g Google Trends oüjmjü qnpb^p^ m^jmjhhp^ h^Jmü ^pm
4 Sh u Doug Laney, "3D Data Management: Controlling Data Volume, Velocity, and Variety", Gartner, file No. 949. 6 February 2001
5 Sh u Performance and Capacity Implications for Big Data,
2014
UpmqntpjntUp mpsmgninrtf t s^jmiUhp^ ushqbtfmU U ^n^n^tfmU shtf^p: Op^Um^ ^nUqmj^U pnpumUhpnrtf mpdhpqphp^ mnp ni. ^m^mnp^ qUhp^ ^n^n^ntpjntUp, ung^mim-^mU tfhq^mjnrtf ^msmp^nq hpm^mpm^nrtf-Uhp^ pmUm^p tf^m^np dmtfmUm^nrtf U mjiU:
PmqtfmqmUnLpjni.Up U^mpmqpnrtf t mjU ^mump, np tfhb s^jmiUhpp qm^u hU 2ms U smpphp mqpjnrpUhp^g, smpphp ^nptfmsUh-pn^, npnUp ^mpnq hU i^Uhi ph ^mnnrg^mb-pmj^U, ph n^ ^mnnig^mbpmj^U: ^mnnrg^mb-pmj^U s^jmiUhpp nrUhUnrtf hU husm^ umh-tfmU^mb ^mnnrg^mbp, frU^p hUmpm^npnrpjnLU t smi^u mpmq npnUnrtf ^msmphi U ^hpinr-bnLpjnrU ^pm^mUmgUhp ^mnnrg^mbpmj^U s^jmiUhpp h^tfUm^mUnrtf ^mh^nrtf hU nhym-g^nU s^jmiUhp^ pmqmjnrtf: H^ ^mnnrg^mb-pmj^U s^jmiUhpp tfUmgmb pninp s^jmiUhpU hU, npnUp ^ntUhU husm^ tfnqhi ^mtf u^htfm (shpusmj^U $mji_hp, ung^mim^mU tfhq^mj^ s^jmiUhp, tlh^spnUmj^U hmqnpqmqpntpjntU-Uhp U mjiU): Smpphp nLunLtfUmu^pnLpjntU-Uhpnttf t, np ^mqtfm^hp^nLpjntUUh-
pnttf s^jmiUhp^ tfns 80-85%-p n^ ^mnntg^mb-pmj^U hU: H^ ^mnntg^m&pmj^U s^jmiUhp^ ^hpintbnLpjmU tf^gngn^ hUmpm^np t pmqhi mpdhpm^np ^U^nptfmg^m, npp ^U^mu-s^ npn2nttfUhp^ ^mjmgtfmUp6:
3V tfnqhip mnmgmp^^hi tp Imq LhjU^^ ^nqtf^g 2001 p^m^mU^U: Ujuop 3V-hp^U tf^m-ghi hU Uu 6-p mpdmUmhm^msntpjntU (Veracity), mpdhp (Value), h^tfUm^np^m&nrp-jnrU (Validity), ^n^n^m^mUntpjnrU (Variability), ^psnrmii (Virtual), ^^qnrm^qmg^m (Visualization)7:
Uh& s^jmiUhpp ^mph^ t oqsmqnp&hi mp-hhusm^mU pmUm^mUnrpjmtfp ods^mb &pmq-php tf2m^hinL hmtfmp, npnUp ^mpnq hU inrbhi pmpq ^Uq^pUhp: Uphhusm^mU pmUm^mUnr-pjmU ntunLtfUmu^pnrpjmU mnmp^mU ^himg^ tfhphUmUhp^ U bpmqphp^ tf2m^nrtfU t8:
Uphhusm^mU pmUm^mUnrpjmU h^tfUm^mU
6 St u Ise Anderson Orobor, 2016. Integration and Analysis of Unstructured Data for Decision Making: Text Analytics Approach
7 Sh 'u Ripon Patgiri, Arif Ahmed. 2016. Big Data: The V's of the Game Changer Paradigm.
8 Sh u Mubashir Hussain, Dr. Manhas, 2016. Artificial intelligence for big data: potential and relevance. International Acadmey of Engineering and Medical Research, Volume-1, ISSUE-1
Регион u Mup, 2019, № 4
JhpnqUhpJg t JhphUmjmlmU nLuntgntJp: Ujq shpJJUp UhpJnibhi t Uppntp UmJnthip 1959 plmlmUJU, Jul $-nJ UJs^hip s^hi t JhphUmjmlmU ntuntgJmU miqnpJpJJ hhsUjmi $np-Jmi umhJmUntJp. «"muhUp, np hmJmlmpq-¿mjJU bpmqJpp unlnpniJ t hiUhinl E ^npAJg, hm2lJ mnUhinl T ^UqJpp U ntUhUmin^ JU^ np P t^hlsJlntpjmU qnpbmlJg, mjU^hu np Upm t$hlsJlnLpjni.Up T ^UqpJ hmJmp, npp ¿m^-IntJ t P t^hlsJlntpjmU qnpbmlgn^, pmph-imllnLJ t E ^npAJ pUpmgpntJ»9' 10:
UhphUmjmlmU ntuntgJmU hJJUmlmU Jh-pnqUhpp ImphiJ t ^Jpm^nph]_ hhsUjmi hplni. qmuhpntJ.
• unLihplJqnpnl ntuntgJmU Jhpnq-Uhp
• mnmUg untihplJqnpJ ntunigJmU JhpnqUhp
UnLihplJqnpnl ntuntgJmU JhpnqUhpp JphUg hhppJU iJUnLJ hU nhqphuJnU U qmum-ImpqJmU hmJmp Um^mshu^mb: fthqphuJnU JhpnqUhpnl ImphiJ t intbhi opJUml mU2mpd qntjpJ qUhpJ ImU^mshuJmU ^UqJpp, Jul qmum^mpqJmU JhpnqUhpnl ImU^mshuhi ^mplmnntJ uUmU^mgntJp:
UnmUg unLihplJqnpJ ntuntgJmU Jhpnq-Uhpp hUmpm^npntpjntU hU smiJu Jpmlm-UmgUhi n^ ImnntglmbpmjJU sljmiUhpJ ihp-inLbnLpjntU: IpmUg JJgngn^ hUmpm^np t sljmiUhpJ Jhg qsUhi opJUm^m^nLpjntUUhp, JpmlmUmgUhi sljmiUhpJ limushpJqmgJm, JUihu UmU JghgUhi sljmiUhpJ ¿m^nqm^m-UntpjntUp IJqntmiJqmgJmjJ U^msmln^:
Hhpp U2^mb JhpnqUhpp Uhp^mjntJu LmjU IJpmnntpjntUUhp ntUhU sUshuntpjmU mjU ninpsUhpntJ, npshq qhUhpmg^ntJ hU Jhb pm-Um^ntpjmJp s^jmiUhp: "hUspnUmlmU pmU-lhpJ hmJmp qpmUp UntjU^hu hhsmppppnt-pjntU hU Uhp^mjmgUntJ, pmUJ np sUshum-lmU 2ms hpUnLjpUhp hUmpm^np t ntuntJUm-uJphi U pmgmsphi Jhbmbmlmi s^jmiUhpJ ntuntJUmuJpJmU JJgngn^, Jul JhphUmjmlmU ntuntgntJp smi^u t qnpbJpUhp mUhpmdh2s opJUm^m^ntpjntUUhpp qsUhint hmJmp: Umu-Um^npm^hu, IhUspnUmlmU pmU^hpntJ Jhb
9 Sh'u Mitchell, T. (1997). Machine Learning. McGraw Hill. p. 2.
10 Sb'u Arthur Samuel,1959. "Some Studies in Machine
Learning Using the Game of Checkers". IBM Journal of
Research and Development. 3 (3): 210-229
s^jmiUhpJ hm^mpmqpntJp U JhphUmjmlmU ntuntgJmU JhpnqUhpJ oqUntpjmJp IhpintbnL-pjntUp Impnq hU hUmpm^npntpjntU smi JpmlmUmgUhi hplpJ JmlpnsUshumlmU gntgm-Uj2UhpJ mnm^hi &?qpJs ImU^mshuntJUhp U imphi m^hiJ mpqjntUm^hs qpmJmlmplmjJU pmqmpm^mUntpjntU: PmgJ mjq IhUspnUmlmU pmUlhpp Jhb sljmiUhpJ IhpinLbnLpjntUp ^mpnq hU ^^pmnhi pmU^hp^ ^pmg^hi^ntp-jntUp U ^^mpntUm^nLpjntUp qUmhmshint, ^n-qhp^ L^mgJmU U ^^mpmhm2^mp^mj^U hmJm-^mpqhpntJ qhqbmpmpnLpjntUUhp^ qhJ ^mj-pmphint hmJmp:
Cus Central Banking hpmsmpm^^ntpjmU U BearingPoint ^nphpqms^m^mU pU^hpntpjmU hmJmshq hhsmqnsntpjmU mpqjntUpUhp^ hhsmqnsntpjmUp JmuUm^gmb m2^mph^ 52 ^hUspnUm^mU pmU^hp^ 56%-p m2^msntJ t Jhb s^jmiUhp Uhpmnnq ^pnjh^s^ ^pm: Zmpg-^mbUhp^ 55%-p Jhb s^jmiUhp^ ^hpinLbntp-jntUp oqsmqnpbntJ t hp^mpmdmJ^hs U ^mp-^mdmJ^hs ^mU^mshuntJUhp^ hmJmp, 42%-p upphu phusm^npJmU, 32%-p ^nqhp^ L^mgJmU qhJ ^mjpmp^, ^u^ 30%-p qhqbmpm-pntpjntUUhp^ pmgmhmjsJmU U ^mU^mshu-JmU hmJmp11:
"T^-J-^iiiWIIMImi*!!
^^■ibiDulll^ npyb pttn Bwjl 1111 f
fcqt^njmjMi^ ii^ut Stfb^mfmjfiiliUf^
%
0 10 20 30 40 50 BO
Q-bmiymmljhp 2. CentralBanking-Ji tynqiffig "P-hpji JJpU mUglmgilmd hmpgJmU mpqniUpUhpp Jhb wijjmLUhpJ JpmnJmU ninpinUhpJ ihpmphpjmi12
^hpUu IhUspnUmlmU pmUlhpJ hmJmp Jhb sljmiUhpJ IhpinLbnLpjmU qi^m^np Jmp-smhpm^hpUhpJg hU hmUqJumUntJ, Um^ U
11 Sh uEmma Glass, Big data in central banks: 2018 survey results
^¿nutqmmlhpp lmqdi[hi tBig data in central banks, 2018 hmpgJmU wiJjmiUhpJ hJJmU ipm
mnmg, qnjntpjntU ntUhgnq s^jmiUhp^ npm^p (hms^m^hu qmpqmgnq hp^pUhp^ hmtfmp), pmU^ np qpmUp hm^m^ tfhb bm^m^ mqtfnt^ hU ^mpntUm^nttf, ^u^ tfjntu ^nqtf^g hu^m-jmbm^mi s^jmiUhp^ tf^m^tfmU qd^mpntpjntUp U hmtfmqmmmu^mh tfmuUmqhsUhp^ Uhp-qpm^nttfp: ^hUspnUm^mU pmU^hpp s^jmiUh-p^ npm^U mnpU^nq ^Uq^pUhpp ^mpnq hU inLbhL mjipUspmUpmj^U mqpjntpUhp^g s^jmi-Uhp^ Ahnpphptfmtfp, ^u^ tfmuUmqhsUhp^ pm-gm^mjntpjntUp ^phUg m2^msm^gUhp^ ^h-pm^mspmustfmU tf^gngn^:
Uju^un^ tfhb s^jmiUhp^ ^hpinLbnLpjntUp hUmpm^np t qmnUnttf tfhphUmjm^mU ntuntg-tfmU tfhpnqUhp^ ^msmphimqnpbtfmU U hm2-^nqm^mU hqnpntpjntUUhp^ m^himgtfmU tf^gn-gn^: ZhsmqnsntpjntUp gntjg t smi^u, np ^hUspnUm^mU pmU^hpp hhsmpppp^mb hU mjq tfhpnqUhpn^ ^phUg mngU qp^mb ^Uq^p-Uhp^ intbtfmtfp: 1m ^mjtfmUm^np^mb t frU^-^hu smpphp mqpjntpUhp^g usmg^nq s^jmi-Uhp^ bm^m^ m^n^, mjU^hu tl hm2^nqm^mU sh^U^mj^ qmpqmgtfmtfp: UmuUm^npm^hu ^hUspnUm^mU pmU^hpnttf tfhpUmjm^mU ntuntgtfmU tfhpnqUhpp ^pmn^nttf hU ^mtf ^mpnq hU ^pmn^hi tfm^pnsUshum^mU gnt-gmUfy?Uhp^ ^mU^mshutfmU, qhqbmpmpntp-jntUUhp^ U ^nqhp^ i^mgtfmU qhtf ^mjpmp^, n^u^hp^ ^mnm^mptfmU U ^pm^mUmg^mb pm-qmpm^mUntpjmU mqqhgntpjmU qUmhmstfmU hmtfmp: Uhb s^jmiUhp^ ^hpinLbnLpjntUp tfhb
hm^mUm^mUntpjmtfp ^mpntUm^ qsU^hi ^hUspnUm^mU pmU^hp^ nt2mqpntpjmU ^hUs-pnUnttf U A-Um^n^hi UpmUg ^nqtf^g ^^UmUum-^mU, sUshum^mU ^hpinLbnLpjntUUhp^ ^pm-^mUmgtfmU U np^nttfUhp^ ^mjmgtfmU tfh^m-U^qtfUhpp:
Oqmmqnpfc^mb qpm^mUntpjmU guU^
1. Demystifying Big Data: A Practical Guide To Transforming The Business of Government (2012)
2. Glass, E. (2018). Big data in central banks: 2018 survey results
3. Halevi, G., Moed H. (2012). The evolution of big data as a research and scientific topic: Overview of the literature
4. Hussain M, Dr. Manhas (2016). Artificial intelligence for big data: potential and relevance. International Acadmey of Engineering and Medical Research, Volume-1, ISSUE-1
5. Jewell, D., Barros, R. D., Diederichs, S., Duijvestijn, L. M., Hammersley, M., Hazra, A. and Zolotow, C. (2014). Performance and Capacity Implications for Big Data, IBM Redbooks
6. Laney, D. (2001, February). 3D Data Management: Controlling Data Volume, Velocity, and Variety, Gartner, file No. 949
7. Mitchell T. (1997). Machine Learning. McGraw Hill. p. 2.
8. Orobor, I. A. (2016). Integration and Analysis of Unstructured Data for Decision Making: Text Analytics Approach
9. Patgiri R., Ahmed A. (2016). Big Data: The V's of the Game Changer Paradigm.
10. Samuel, A. (1959). Some Studies in Machine Learning Using the Game of Checkers. IBM Journal of Research and Development, 3 (3), 210-229
