さくらインターネット研究所で研究に再挑戦した私の半年間の取り組み

Transcript

1. ͘͞ΒΠϯλʔωοτגࣜձࣾ
   (C) Copyright 1996-2019 SAKURA Internet Inc
   ͘͞ΒΠϯλʔωοτݚڀॴ
   ͘͞ΒΠϯλʔωοτݚڀॴͰݚڀʹ࠶௅ઓͨ͠
   ࢲͷ൒೥ؒͷऔΓ૊Έ
   2020/01/16 ݚڀһ ௽ా തจ
   ͘͞Βͷ༦΂ ݚڀॴφΠτ
   

   View Slide

2. 2
   1. ࣗݾ঺հ
   2. ݚڀॴͰͷ೔ৗ
   3. ͜Ε·Ͱͷݚڀͱࠓޙͷల๬
   4. ·ͱΊ
   ໨࣍
   

   View Slide

3. 1.
   ࣗݾ঺հ
   

   View Slide

4. 4
   ࣗݾ঺հ
   ௽ా തจʢͭΔͨ ͻΖ;Έʣ
   ɹɹॴଐɹ͘͞ΒΠϯλʔωοτݚڀॴ ݚڀһʢ2019೥8݄ʙʣ
   ɹɹֶྺɹ۝भେֶେֶӃ ࡐྉ෺ੑ޻ֶઐ߈ म࢜՝ఔमྃ
   ɹɹɹɹɹ۝भେֶେֶӃ ࡐྉ෺ੑ޻ֶઐ߈ ത࢜՝ఔதୀ
   ɹɹܦྺɹࡐྉ޻ֶͷݚڀɼফ๷࢜ɼػցֶशΤϯδχΞɼ
   ɹɹɹɹɹΠϯϑϥΤϯδχΞΛܦͯɼݱ৬
   ڵຯྖҬɹػցֶशͱྔࢠίϯϐϡʔλɾΞχʔϦϯά
   @tsurubee3
   https://blog.tsurubee.tech/
   

   View Slide

5. 5
   ܦྺ·ͱΊ
   ܦྺ ظؒ
   ࡐྉ޻ֶͷݚڀऀ 4೥
   ফ๷࢜ 3೥
   ITΤϯδχΞ 3೥
   ৘ใ޻ֶͷݚڀऀ 6ϲ݄ ←ΠϚίί
   

   View Slide

6. 2.
   ݚڀॴͰͷ೔ৗ
   

   View Slide

7. • ຖ೔ఆ࣌લ45෼ؒ
   • ೚ҙࢀՃ
   • ຊ౰ʹࡶஊ͚ͩͷͱ͖΋͋Ε͹ɼ
   ݚڀͷٞ࿦Λ͢Δͱ͖΋͋Δ
   7
   • ֤ݚڀһ͕ڵຯɾؔ৺ʹ೚ͤͯɼ໘ന͍ͱࢥ͏ςʔϚʹͲ͠Ͳ͠औΓ૊ΜͰ͍͘
   • ֤ݚڀһ͕ίϯηϓτͰඳ͍ͨੈքΛ࣮ݱ͢Δ্Ͱඞཁͳ՝୊Λࣗ཯తʹݟ͚ͭɼ
   ݚڀΛਐΊ͍ͯ͘
   ݚڀॴͷελΠϧɿࣗ཯ɾ෼ࢄɾڠௐ
   ࣗ཯
   • ݚڀһ͸஍ཧతʹ෼ࢄ͍ͯ͠Δ
   ౦ژ6໊ʢ٬һݚڀһ2໊ʣɾେࡕ2໊ɾ෱Ԭ2໊ʢ2020೥1݄ݱࡏʣ
   • िҰճ։࠵
   • ࿩͍ͨ͠ਓ͕࿩͢ελΠϧ
   • ൃදͷࣄલใࠂɾࣄޙใࠂɼ
   ݚڀͷΞΠσΞͳͲ༷ʑ
   • ෆఆظ
   • ݸผʹೋਓͰ࿩͢ͳͲ
   • ΞΠσΞΛฉ͍ͯ΋ΒͬͨΓɼ
   ࠷ۙ΍͍ͬͯΔ͜ͱΛ࿩ͨ͠Γ
   ఆྫձ ࡶஊλΠϜ ͦͷଞ
   ෼ࢄ
   ڠௐ
   

   View Slide

8. 8
   ݸਓͷελΠϧ
   ಇ͖ํ
   • ࿦จࣥච
   • ൃදɾσΟεΧογϣϯͷࢿྉ࡞੒
   • αʔϕΠɼͳͲ
   ࿦จࣥචɿ೔ຊޠ1ɼӳޠ1
   ࢿྉ࡞੒ɿֶձ1ɼษڧձ3ɼاۀσΟεΧογϣϯ2
   ൒೥ؒ
   ࠓޙ
   ࣗ෼ͷઐ໳෼໺ͷཱ֬ͱͦͷ෼໺Ͱͷത࢜߸औಘΛ໨ࢦ͢
   ۀ຿
   • बۀ࣌ؒɿ9:30ʙ18:30ʢ10෼୯ҐͰεϥΠυՄʣ
   • िʹ2ɼ3ճఔ౓ϦϞʔτϫʔΫʢࡏ୐ۈ຿ɿ8:30ʙ17:30ʣ
   • ݄ʹ1ɼ2ճఔ౓ग़ு
   • ࠃ಺ɾࠃࡍֶձʢൃදͷ༗ແʹؔΘΒͣʣ
   • اۀͱͷσΟεΧογϣϯ
   • ݚڀॴ߹॓ɼͳͲ
   

   View Slide

9. 3.
   ͜Ε·Ͱͷݚڀͱࠓޙͷల๬
   

   View Slide

10. 10
    ݱࡏͷݚڀςʔϚ
    ҎԼͷ2ͭΛ࣠ʹݚڀ׆ಈΛߦ͍ͬͯΔ
    ̍ɽϢʔβʹมߋΛཁٻͤͣʹγεςϜมԽʹ௥ैՄೳͳ
    ɹɹSSHϓϩΩγαʔό
    ̎ɽྔࢠίϯϐϡʔλɾΞχʔϦϯάϚγϯΛ༻͍ͨݚڀߏ૝
    

    View Slide

11. 11
    ݱࡏͷݚڀςʔϚ
    ҎԼͷ2ͭΛ࣠ʹݚڀ׆ಈΛߦ͍ͬͯΔ
    ̍ɽϢʔβʹมߋΛཁٻͤͣʹγεςϜมԽʹ௥ैՄೳͳ
    ɹɹSSHϓϩΩγαʔό
    ̎ɽྔࢠίϯϐϡʔλɾΞχʔϦϯάϚγϯΛ༻͍ͨݚڀߏ૝
    

    View Slide

12. 12
    ssh username@
    SSH
    Client
    • WebαʔϏεΛࢧ͑ΔΠϯϑϥ͸ɼར༻ऀ͔Βͷଟ༷ͳཁٻ΍؀ڥͷมԽ౳ʹԠͯ͡ɼ
    ਝ଎͔ͭॊೈʹγεςϜߏ੒Λมߋ͢Δ͜ͱ͕ٻΊΒΕΔɽ
    • αʔϏε΁ͷଟ༷ͳཁٻʹԠͯ͡γεςϜߏ੒Λਝ଎ʹมߋ͍ͯ͘͜͠ͱ͕ٻΊΒΕΔ
    ঢ়گʹ͓͍ͯ͸ɼγεςϜͷӡ༻؅ཧ΋มߋʹ௥ैͰ͖Δඞཁ͕͋Δɽ
    • Ұํɼ҆શͳϦϞʔτ઀ଓαʔϏεͱͯ͠αʔό؅ཧʹ޿͘ར༻͞Ε͍ͯΔSSH͸ɼ
    Ϣʔβ͕ར༻͢ΔαʔόͷIPΞυϨε·ͨ͸ϗετ໊Λࢦఆͯ͠઀ଓཁٻΛૹΔ࢓૊Έ
    Ͱ͋Δɽ
    എܠɿγεςϜมԽʹ௥ैͰ͖Δӡ༻؅ཧ
    Ϣʔβ มߋ
    Server
    Server
    αʔόͷIPΞυϨεɾ
    ϗετ໊ͷมߋͳͲ
    Ϣʔβ͕มߋޙͷ৘ใ
    Λ஌Δඞཁ͕͋Δ
    

    View Slide

13. 13
    sshr: SSHϓϩΩγαʔό
    γεςϜ؅ཧऀ͕ࣗ༝ʹ૊ΈࠐΈՄೳͳϑοΫؔ਺Λ༻͍ͯγεςϜมԽʹ௥ैͰ͖Δ
    sshr※1ͱ͍͏SSHϓϩΩγαʔόΛఏҊ
    SSHΫϥΠΞϯτ
    ssh username@hostname
    Ϣʔβ໊ ઀ଓઌϗετ
    ؅ཧσʔλ
    ϑοΫؔ਺
    SSH
    ϓϩΩγαʔό
    αʔό܈
    ※1 https://github.com/tsurubee/sshr
    • Ϣʔβʹ༻͍ΔΫϥΠΞϯτπʔϧͷ੍ݶ΍มߋΛ՝͞ͳ͍
    • ૊ΈࠐΉϑοΫؔ਺ͷΈͷमਖ਼ͰϓϩΩγαʔόͷಈ࡞Λࣗ༝ʹม͑ΒΕΔͨΊɼ
    γεςϜͷ࢓༷มߋʹରͯ͠ߴ͍֦ுੑΛ༗͢Δ
    

    View Slide

14. 14
    IOTS2019Ͱൃද͠·ͨ͠
    2019೥12݄5ʙ6 ೔ʹ։࠵͞Εͨୈ12ճΠϯλʔωοτͱӡ༻ٕज़
    γϯϙδ΢Ϝ (IOTS2019)Ͱൃද͠·ͨ͠
    

    View Slide

15. 15
    ࠓޙͷల๬
    ΫϥΠΞϯτ͔ΒݟΔͱɼSSHͷϢʔβ໊ʹඥ͍ͮͨ1ͭͷαʔόʹ͔͠SSH઀ଓͰ͖ͣɼ
    ෳ਺ͷαʔόͷத͔Βҙਤతʹ઀ଓઌΛࢦఆͯ͠઀ଓཁٻΛૹΔ͜ͱ͕Ͱ͖ͳ͍
    SSHΫϥΠΞϯτ
    ᶃ઀ଓཁٻ
    ᶄSSHϢʔβ໊ ᶅαʔόߏ੒৘ใ
    ؅ཧσʔλ
    ϑοΫؔ਺
    SSH
    ϓϩΩγαʔό
    ղܾҊ
    ՝୊
    ɾκʔϯ΍ϩʔϧͳͲͷλά৘ใΛ΋ͱʹબ୒͢Δ
    ɾো֐ঢ়گɾαʔόෛՙͳͲͰιʔτͯ͠ϑΟϧλϦϯά
    ଟछଟ༷ͳϩʔϧ΍σόΠε͕͋ΓɼͦΕΒ͕࣌ʑࠁʑͱมԽ͢ΔϦιʔεͷ؅ཧʹ͓͍ͯ
    ΫϥΠΞϯτ͕ͦΕΒͷঢ়گΛஞҰ೺Ѳ͢Δ͜ͱͳ͘ɼ໨తαʔόΛݕࡧͯ͠઀ଓͰ͖Δ࢓૊Έ
    ᶆαʔόߏ੒৘ใ
    ᶇ઀ଓઌαʔόΛબ୒
    ᶈ઀ଓཁٻ
    $ ssh tsurubee@sshr
    +-----+----------+---------+--------+-----------+----------+
    | No. | Hostname | Zone | Role | CPU usage | Status |
    +----------------+---------+--------------------+----------+
    | 1 | server01 | tokyo | db | 90 | critical |
    | 2 | raspi01 | fukuoka | sensor | 70 | warning |
    ɿɹɹɹɹ ɹ ɿ
    

    View Slide

16. 16
    ݱࡏͷݚڀςʔϚ
    ҎԼͷ2ͭΛ࣠ʹݚڀ׆ಈΛߦ͍ͬͯΔ
    ̍ɽϢʔβʹมߋΛཁٻͤͣʹγεςϜมԽʹ௥ैՄೳͳ
    ɹɹSSHϓϩΩγαʔό
    ̎ɽྔࢠίϯϐϡʔλɾΞχʔϦϯάϚγϯΛ༻͍ͨݚڀߏ૝
    

    View Slide

17. 17
    ྔࢠίϯϐϡʔλͱ͸
    ྔࢠίϯϐϡʔλͱ͸ɼྔࢠྗֶͷݪཧʹج͍ͮͨ৽͍͠ίϯϐϡʔλͷ֓೦Ͱ͋Δɽ
    ྔࢠίϯϐϡʔλͷΞΠσΞ͸ɼ1982೥ͷFeynmanͷ
    ࿦จ͕࢝·Γͱ͞Ε͍ͯΔɽ
    ैདྷίϯϐϡʔλͱͷҧ͍ɿ৘ใͷද͠ํ
    ϒϩοϗٿ※1
    ※1 https://ja.wikipedia.org/wiki/%E3%83%96%E3%83%AD%E3%83%83%E3%83%9B%E7%90%83
    ৘ใͷجຊ୯ҐɿϏοτ ྔࢠϏοτ
    ྔࢠྗֶ
    or
    0
    1
    0
    1
    0·ͨ͸1ͲͪΒ͔
    ͷঢ়ଶΛͱΔ
    0ͱ1ͷ྆ํͷঢ়ଶ
    Λಉ࣌ʹͱΔ
    ʢॏͶ߹Θͤঢ়ଶʣ
    

    View Slide

18. 18
    ྔࢠίϯϐϡʔλɾΞχʔϦϯάϚγϯ
    ྔࢠίϯϐϡʔλ(ήʔτํࣜ) ΞχʔϦϯάϚγϯ(ΠδϯάϚγϯ)
    Google IBM Microsoft IonQ
    σδλϧճ࿏
    ௒ిಋճ࿏
    ෋࢜௨
    ೔ཱ
    D-Wave NEC
    ྔࢠྗֶͷݪཧΛར༻
    

    View Slide

19. 19
    ྔࢠήʔτϚγϯͷݱঢ়
    ྔࢠϏοτ਺ͷ֦େ
    ྔࢠ௒ӽੑͷ࣮ূ
    Ϗοτ਺΋೥ʑ৳ͼ͍ͯΔ͕ɼ࣮༻తͳ໰୊
    Λղ͘ʹ͸ɼࠓޙͷݚڀ։ൃ͕଴ͨΕΔɽ
    ग़యɿʰ1೥Ͱूੵ౓͕ڻҟతʹ޲্ͨ͠ྔࢠίϯϐϡʔλʱ
    https://jbpress.ismedia.jp/articles/-/54979
    Google͕2019೥10݄ʹɼྔࢠίϯϐϡʔλͷܭࢉೳྗ͕ɼεʔύʔίϯϐϡʔλͳͲ
    ैདྷܕͷίϯϐϡʔλΛ্ճΔ͜ͱΛࣔ͢ʮྔࢠ௒ӽੑʯΛ࣮ূͨ͠ͱൃද※1
    ࠷ઌ୺ͷεʔύʔίϯϐϡʔλͰ͸ղ͘ͷʹ໿1ສ೥͔͔ΔܭࢉΛGoogleͷྔࢠίϯ
    ϐϡʔλ͸200ඵͰղ͍ͨͱ͞Ε͍ͯΔɽ(໿16ԯഒ)
    ※1 F. Arute et al.: Quantum supremacy using a programmable superconducting processor, Nature, Vol. 574, 505-510 2019.
    

    View Slide

20. 20
    (ྔࢠ)ΞχʔϦϯάϚγϯͷݱঢ়
    D-Waveͷ঎༻ԽͱԠ༻ࣄྫͷ޿͕Γ
    ࠃ಺اۀͷ࣮ূ࣮ݧ
    2011೥ʹ঎༻ΞχʔϦϯάϚγϯD-Waveͷొ৔
    ϑΥϧΫεϫʔήϯ͕๺ژͷࢢ಺ͷަ௨ौ଺ΛD-WaveΛ
    ׆༻ͯ͠ղফ͢Δࣾձ࣮ݧΛߦͬͨ※1
    ※1 F. Neukart et al.: Traffic flow optimization using a quantum annealer, Frontiers in ICT, Vol. 4, 1-6 2017.
    ೔ཱ੡࡞ॴͷCMOSΞχʔϦϯάϚγϯ΍෋࢜௨ͷ
    σδλϧΞχʔϥ౳Λ׆༻࣮ͨ͠ূ࣮ݧ͕ߦΘΕͯ
    ͍Δ
    https://www.hitachi.co.jp/New/cnews/month/2020/01/0108.html
    

    View Slide

21. 21
    ྔࢠίϯϐϡʔλɾΞχʔϦϯάϚγϯ
    ྔࢠίϯϐϡʔλ(ήʔτํࣜ) ΞχʔϦϯάϚγϯ(ΠδϯάϚγϯ)
    Google IBM Microsoft IonQ
    σδλϧճ࿏
    ௒ిಋճ࿏
    ෋࢜௨
    ೔ཱ
    D-Wave NEC
    ྔࢠྗֶͷݪཧΛར༻
    

    View Slide

22. 22
    ͜Ε·Ͱͷ͘͞ΒͷऔΓ૊Έ
    https://www.sakura.ad.jp/information/pressreleases/2018/10/18/1968198517/
    1೥Ҏ্લ͔Β೔ཱ੡࡞ॴ͕։ൃͨ͠
    ΞχʔϦϯάϚγϯͷධՁΛߦͳ͖ͬͯͨ
    ʢ౰ݚڀॴ٠஍͕୲౰ʣ
    

    View Slide

23. 23
    ૊߹ͤ࠷దԽ໰୊
    ཭ࢄଟม਺ؔ਺ͷ࠷খ஋ʢ͋Δ͍͸࠷େ஋ʣ͓Αͼͦͷ࠷খ஋Λ༩͑Δม਺ͷ
    ૊Έ߹ͤΛٻΊΔ໰୊
    ྫͱͯ͠ɼ८ճηʔϧεϚϯ໰୊ɼφοϓαοΫ໰୊ɼδϣϒγϣοϓ
    εέδϡʔϦϯά໰୊ͳͲ͕ڍ͛ΒΕɼଟ༷ͳ෼໺ʹ಺ࡏ͍ͯ͠Δɽ
    ଟ͘ͷ૊߹ͤ࠷దԽ໰୊͸ैདྷͷίϯ
    ϐϡʔλͰ͸ޮ཰తʹղ͘͜ͱ͕ࠔ೉Ͱ
    ͋Γɼ໰୊ͷղۭؒͷ୳ࡧ͕ඞཁͰ͋Δ
    ૊߹ͤ࠷దԽ໰୊ͱ͸
    ղ͘ͷ͕೉͍͠ʁ ղͷީิ਺ʢʹܭࢉ࣌ؒʣ
    ࢦ਺ؔ਺త૿Ճ
    ʹ૊߹ͤരൃ
    ໰୊ͷαΠζ
    ϝλώϡʔϦςΟοΫ
    ΞϧΰϦζϜ
    

    View Slide

24. 24
    ϝλώϡʔϦεςΟοΫ
    ϝλώϡʔϦεςΟοΫͱ͸
    ಛఆͷ໰୊ͷΈΛର৅ͱ͢ΔͷͰ͸ͳ͘ɺ༷ʑͳ໰୊ʹରͯ͠ɺ
    ൺֱత୹࣌ؒͰۙࣅղΛޮ཰Α͘ٻΊΔղ๏
    ΞϧΰϦζϜͷྫ
    • Ҩ఻తΞϧΰϦζϜ
    • ٜίϩχʔ࠷దԽ
    • ཻࢠ܈࠷దԽ
    • ྔࢠΞχʔϦϯά
    ਐԽܭࢉΞϧΰϦζϜ
    Ϋϥ΢υίϯϐϡʔςΟϯά΍
    ΤοδɾϑΥάίϯϐϡʔςΟ
    ϯάͷจ຺Ͱͷݚڀࣄྫ΋͋Δ
    https://www.sciencedirect.com/science/article/pii/S1110866515000353
    https://dl.acm.org/doi/10.1145/3287921.3287984
    

    View Slide

25. 25
    ྔࢠΞχʔϦϯά
    • ྔࢠΞχʔϦϯάͱ͸ɼྔࢠྗֶͷ๏ଇΛར༻ͯ͠ɼ
    ͋Δछͷ৘ใॲཧΛ͢ΔͨΊͷ࿮૊ΈͰ͋Δ
    ྔࢠΞχʔϦϯάͱ͸
    • ૊߹ͤ࠷దԽ໰୊ʹର͢ΔϝλώϡʔϦεςΟοΫ
    ͳղ๏ͱͯ͠஫໨͞Ε͍ͯΔ
    ̍ɽେن໛ͳ૊Έ߹Θͤ࠷దԽ໰୊Λߴ଎ʹղ͘͜ͱ͕ظ଴͞ΕΔ※2
    ̎ɽجఈঢ়ଶ͕ॖୀ͍ͯ͠Δ৔߹ʹɼภͬͨղ͕ಘΒΕΔ※3ʢΞϯϑΣΞαϯϓϦϯάʣ
    ྔࢠΞχʔϦϯάͷಛ௃
    ※1 T. Kadowaki and H. Nishimori: Quantum annealing in the transverse Ising model, Phys. Rev. E, Vol. 58, 5355 1998.
    ※2 V. S. Denchev et al.: What is the Computational Value of Finite Range Tunneling?, Phys. Rev. X, Vol. 6, 031015 2016.
    ※3 B. H. Zhang et al.: Advantages of Unfair Quantum Ground-State Sampling, Scientific Reports, Vol. 7, 1044 2017.
    ྔࢠΞχʔϦϯά͸1998೥ͷ࿦จͰ೔ຊਓ
    ͕ཧ࿦ΛߟҊͨ͠※1
    • ΠδϯάϞσϧͷجఈঢ়ଶΛྔࢠྗֶతͳΏΒ͗Λ
    ར༻ͯ͠୳ࡧ͢Δ͜ͱͰ૊߹ͤ࠷దԽ໰୊Λղ͘
    ΠδϯάϞσϧ
    

    View Slide

26. 26
    1. ߴ଎ͳ૊Έ߹Θͤ࠷దԽͷղ๏
    ※1 V. S. Denchev et al.: What is the Computational Value of Finite Range Tunneling?, Phys. Rev. X, Vol. 6, 031015 2016.
    ※2 ੢৿ लູ, େؔ ਅ೭: ྔࢠΞχʔϦϯάͷجૅ, ڞཱग़൛, 2018.
    ૊Έ߹Θͤͷ਺͕๲େͰैདྷͷख๏Ͱ͸࠷దղͷ୳ࡧʹ͕͔͔࣌ؒΔ໰୊Λ࣮༻తͳ
    ࣌ؒ಺Ͱղ͘͜ͱ͕ظ଴Ͱ͖Δɽ
    γϛϡϨʔςΟουΞχʔϦϯάʢSAʣʹൺ΂ͯ
    ྔࢠϞϯςΧϧϩ๏ʢQMCʣɼD-Wave͸໰୊αΠζͷ
    ֦େʹର͢Δ܏͖͕খ͍͞ɽ
    ࠷దԽ୳ࡧ·Ͱʹ͔͔Δ࣌ؒͱ໰୊ͷαΠζͷؔ܎※1
    ҎԼͷΑ͏ͳٞ࿦΋͋Δ※2
    ɾͦ΋ͦ΋SA͸ϝλώϡʔϦεςΟοΫͳΞϧΰϦζϜ
    ɹͰ͋Γɼߴ଎Ͱ͸ͳ͍ɽ
    ɾGPUͷ׆༻΍ฒྻॲཧͰSAɼQMC΋ߴ଎ԽͰ͖Δɽ
    

    View Slide

27. 27
    2. ΞϯϑΣΞαϯϓϦϯά
    ϑϦʔεϐϯͷ਺͕ଟ͍εϐϯ഑ஔ͕બ୒తʹ
    ಘΒΕ΍͍͢※1
    ※1 Y. Matsuda et al.: Ground-state statistics from annealing algorithms: quantum versus classical approaches, New Journal of Physics, Vol. 11, 073021 2009.
    ※2 S. Mandrà et al.: Exponentially-Biased Ground-State Sampling of Quantum Annealing Machines with Transverse-Field Driving Hamiltonians, Phys. Rev. Lett.,
    Vol. 118, 070502 2017.
    ※3 D. Bertsimas et al.: Robust optimization with simulated annealing, Journal of Global Optimization, Vol. 48, 323-334 2009.
    جఈঢ়ଶ͕ॖୀ͍ͯ͠Δ৔߹ɼऔΓ͏Δશͯͷجఈঢ়ଶ͕౳͍֬͠཰ͰಘΒΕͳ͍͜ͱ͕
    ྔࢠϞϯςΧϧϩ๏※1΍D-Wave※2Ͱࣔ͞Ε͍ͯΔʢSAͰ͸౳͍֬͠཰ͰಘΒΕΔʣ
    QA͕ಛఆͷجఈঢ়ଶʹ౸ୡ͢Δ૬ରස౓ͷώετάϥϜ
    ղͷ૊Έ׵͑΍͢͞ʹܨ͛ΒΕͳ͍ͩΖ͏͔ɽ
    কདྷͷෆ֬ఆΛड͚ೖΕͨϩόετ࠷దԽ※3
    

    View Slide

28. 28
    ITΠϯϑϥͱ࠷దԽ
    ෳࡶੑ͕૿͢ITΠϯϑϥ
    ͳͲ༷ʑͳཁҼ͕བྷΈ߹͍ɼγεςϜΛࢧ͑ΔΠϯϑϥ͸ෳࡶԽ͍ͯ͠Δɽ
    • ίϯςφͳͲͷԾ૝Խٕज़ͷීٴ
    • ෼ࢄγεςϜɾϚΠΫϩαʔϏεԽ
    • IoTσόΠεͷීٴʢΤοδɾϑΥάίϯϐϡʔςΟϯάʣ
    ITΠϯϑϥͷ࠷దԽ
    • ΠϯϑϥͷෳࡶԽʹ൐͍ɼϦιʔε؅ཧɼ؂ࢹɾҟৗݕ஌ɼίετ࡟ݮͳͲͷ໰୊΋ෳࡶԽ
    ͍ͯ͠Δɽ
    • ਓ͕ؒ༩͑ͨϧʔϧϕʔεͰͷ؅ཧ͸ݶք͕͋ΔͨΊɼಈతʹมԽ͢Δঢ়گΛߟྀ͠ͳ͕Β
    ࠷దԽ͢Δ࢓૊ΈΛ࣮ݱ͍ͨ͠ɽ
    Ϋϥ΢υίϯϐϡʔςΟϯά΍ΤοδɾϑΥάίϯϐϡʔςΟϯάͳͲզʑͷ෼໺Ͱ
    ղ͖͍ͨ૊߹ͤ࠷దԽ໰୊Λߟ͍͑ͯ͘
    

    View Slide

29. 29
    1. φοϓαοΫ໰୊
    ※1 C. Guerrero et al.: Genetic Algorithm for Multi-Objective Optimization of Container Allocation in Cloud Architecture, Journal of Grid Computing, Vol. 16, 113-135 2018.
    ※2 Y. Gao et al.: A multi-objective ant colony system algorithm for virtual machine placement in cloud computing, J. Comput. Syst., Vol. 79, 1230-1242 2013.
    Physical Node
    ෳ਺ͷ෺ཧϊʔυʹରͯ͠ίϯςφɾԾ૝ϚγϯʢVMʣΛͲͷΑ͏ʹ഑ஔ͢Δͷ͔
    ໰୊
    ίϯςφɾVMͷऩ༰ઃܭ͸ɺγεςϜશମͷύϑΥʔϚϯε΍৴པੑɼεέʔϥϏϦςΟʹେ͖ͳӨڹ
    Λ༩͑Δɽ͜Ε·ͰʹҨ఻తΞϧΰϦζϜ※1΍ٜίϩχʔ࠷దԽ※2Ͱ࠷దԽͨ͠ࣄྫͳͲ͕ଘࡏ͢Δɽ
    Container
    or VM
    

    View Slide

30. 30
    2. δϣϒγϣοϓεέδϡʔϦϯά໰୊
    ஍ཧతʹ෼ࢄͨ͠ίϯϐϡʔςΟϯάϦιʔεʹͲ͏λεΫΛׂΓ౰ͯΔ͔
    ໰୊
    ҟͳΔੑೳΛ༗͢Δෳ਺ͷϊʔυʹෳ਺ͷλεΫΛޮ཰తʹׂΓ౰ͯΔ͜ͱͰɼશλεΫͷ׬ྃ࣌ؒΛ
    ୹ॖ͢ΔɽλεΫಉ࢜ʹґଘؔ܎΍༏ઌॱҐ͕͋Δ৔߹΋૝ఆ͞ΕΔɽ
    ※1 B. M. Nguyen et al.: Evolutionary Algorithms to Optimize Task Scheduling Problem for the IoT Based Bag-of-Tasks Application in Cloud–Fog Computing Environment,
    Appl. Sci., Vol. 9, 1730 2019.
    ※1
    https://jp.alibabacloud.com/about/what-is-edge-computing
    

    View Slide

31. 31
    3. ΫϥελϦϯά
    ίϯϐϡʔςΟϯάϦιʔεΛෛՙঢ়گ౳ͰΫϥελϦϯά͢Δɽ
    ૬ରతʹҟৗͳৼΔ෣͍Λ͍ͯ͠Δ΋ͷΛݕग़͢Δɽ
    ໰୊
    ※1 B. Taskar et al.: Probabilistic Classification and Clustering in Relational Data, the Seventeenth International Joint Conference on Artificial Intelligence, 4-10 2001.
    ࣌ؒ
    ͭͷ఺ʹ୆ͷαʔόɼ7.ɼ*P5
    σόΠε౳ͷଟ࣍ݩ࣌ܥྻσʔλͷ
    ৘ใ͕ೖ͍ͬͯΔΠϝʔδ
    https://towardsdatascience.com/semantic-similarity-classifier-and-
    clustering-sentences-based-on-semantic-similarity-a5a564e22304
    $16ɾϝϞϦར༻཰
    ౳ͷϝτϦΫε
    ͦΕͧΕΛಠཱͯ͠औΓѻΘͣޓ͍ʹ૬ؔΛ࣋ͭ
    ؔ܎σʔλͱͯ͠ΫϥελϦϯά͢Δख๏※1΋༗ޮ
    Ͱ͋Δͱߟ͑Δɽ
    

    View Slide

32. 32
    ࠓޙͷల๬
    ॴ๬ͷ࣌ؒεέʔϧͰղ͚Δ͔ʁ
    ໰୊ઃఆɿΫϥ΢υίϯϐϡʔςΟϯάɼΤοδɾϑΥάίϯϐϡʔςΟϯάʹ͓͍ͯղ͖͍ͨ
    ɹɹɹɹɹ࠷దԽ໰୊Λఆٛ͢Δ
    ఆࣜԽɿ໨తؔ਺ͷఆࣜԽ΍ΠδϯάϞσϧ΁ͷམͱ͠ࠐΈ
    ධՁɿ֤छղ๏Λ༻͍ͯ໰୊Λղ͘
    https://link.springer.com/content/pdf/10.1007/s10898-009-9496-x.pdf
    ग़͖ͯͨղͷ෼෍ɾੑ࣭͸Ͳ͏͔ʁ
    ղͷੑ࣭͕ύϥϝʔλมಈʹରͯ͠ؤڧͰ͋Γɼܥͷ҆ఆੑΛҡ࣋͠
    ͨ··༰қʹ૊Έ׵͑Մೳͳϩόετੑ΋ͭ࠷దԽ͕࣮ݱͰ͖Ε͹ɼ
    ಈతʹมԽ͢ΔγεςϜ؀ڥʹ͓͍ͯɼকདྷͷෆ֬ఆੑΛड͚ೖΕͭ
    ͭܧଓతʹ࠷దԽՄೳͳ࢓૊Έ͕࣮ݱͰ͖ΔͷͰ͸ͳ͍͔
    

    View Slide

33. 33
    ·ͱΊ
    • ͘͞ΒΠϯλʔωοτݚڀॴͰ6೥ͿΓʹݚڀʹ௅ઓ͠·ͨ͠ɽ
    • ࣗ཯෼ࢄڠௐతͳݚڀॴͰշదͳݚڀϥΠϑΛૹ͍ͬͯ·͢ɽ
    • ࠓޙ͸ɼSSHؔ࿈͓ΑͼྔࢠίϯϐϡʔλɾΞχʔϦϯάؔ࿈ͷݚڀΛ
    ਐΊ͍͖ͯ·͢ɽ
    • ·ͨࠓޙͷݚڀ੒ՌΛͲ͔͜Ͱ͓࿩͠͠·͢ʂ
    

    View Slide

34. 34
    ँࣙ
    ຊൃදͰ঺հͨ͠sshrͷ։ൃΛਐΊΔʹ͋ͨΓɺଟେͳΔ͝ࢧԉͱ
    ͝ॿݴΛࣀΓ·ͨ͠GMOϖύϘגࣜձࣾͷϗεςΟϯάࣄۀ෦ͷ
    օ༷Λ͸͡Ίଟ͘ͷํʑʹް͘ײँΛਃ্͛͠·͢ɽ
    

    View Slide