$30 off During Our Annual Pro Sale. View Details »

クックパッドと機械学習（短縮版）/MLTBP

j.harashima
March 28, 2019
0
4k

 クックパッドと機械学習（短縮版）/MLTBP

j.harashima

March 28, 2019
Tweet

More Decks by j.harashima

See All by j.harashima
日本語レシピデータセットの継続的な構築と複合的な利用/JED2022
junharashima
0
2.4k
自然言語処理の基礎と応用 〜 料理と医療を題材として 〜 /JADI2021
junharashima
0
1.6k
企業での研究開発の楽しさと苦労/WAP-Tech-Talk
junharashima
1
6.4k
クックパッドにおける研究開発/HCG2020
junharashima
0
920
Calorie Estimation in a Real-World Recipe Service/iaai-20
junharashima
0
10k
クックパッドと NLP/CV/nlpaper-challenge
junharashima
0
64
AI 部門の実戦投入/repro-tech-meetup
junharashima
0
910
クックパッドと機械学習/uec-aix-seminar
junharashima
0
6.9k
実データにふれる自然言語処理インターンシップ/internship2018
junharashima
1
12k

Featured

See All Featured
Teambox: Starting and Learning
jrom
124
7.8k
Fight the Zombie Pattern Library - RWD Summit 2016
marcelosomers
226
16k
Imperfection Machines: The Place of Print at Facebook
scottboms
254
12k
Intergalactic Javascript Robots from Outer Space
tanoku
260
25k
Helping Users Find Their Own Way: Creating Modern Search Experiences
danielanewman
8
1.2k
Three Pipe Problems
jasonvnalue
89
8.8k
Dealing with People You Can't Stand - Big Design 2015
cassininazir
351
21k
Visualizing Your Data: Incorporating Mongo into Loggly Infrastructure
mongodb
32
6k
Design by the Numbers
sachag
271
17k
Rails Girls Zürich Keynote
gr2m
87
12k
WebSockets: Embracing the real-time Web
robhawkes
58
5.9k
Thoughts on Productivity
jonyablonski
48
2.6k

Transcript

  1. .BDIJOF-FBSOJOH5FBN#VJMEJOH1JUDI ΫοΫύουͱػցֶशɹ ΫοΫύουגࣜձࣾݪౡ७ ʢ୹ॖ൛ʣ

  2. ೥݄ ೥݄ ژ౎େֶࠇڮɾՏݪݚڀࣨ഑ଐ ΫοΫύουೖࣾɺαʔϏε։ൃ෦໳഑ଐ ത࢜ʢ৘ใֶʣऔಘ ΤϯδχΞʢ3VCZPO3BJMTػցֶशʣ ೥݄ Ϧαʔνϟʔʢࣗવݴޠॲཧ৘ใݕࡧʣ ݪౡ७ ೥݄

    Ϛωʔδϟʔʢਓࣄ޿ใࡒ຿ʜʣ ݚڀ։ൃ෦໳৽ઃɺಉ෦໳഑ଐ

  3. ΫοΫύου Πϯλʔωοτ্ͰϨγϐͷ౤ߘɾݕࡧ౳͕Ͱ͖Δ ೔ຊ࠷େͷྉཧϨγϐαʔϏε

  4. Ԋֵ ೥݄ ༗ݶձࣾίΠϯʢݱΫοΫύουגࣜձࣾʣઃཱ ೥݄ ೥݄ ೥݄ ೥݄ ೥݄ ೥݄ Ϩγϐͷ౤ߘɾݕࡧαʔϏεLJUDIFO!DPJO։࢝

    ΫοΫύουʹαʔϏε໊มߋ ೥݄ ϓϨϛΞϜαʔϏε։࢝ Ϛβʔζ্৔ ౦ূҰ෦ʹࢢ৔มߋ ೥݄ ւ֎ల։Λຊ֨Խ ઃཱप೥ ϓϨϛΞϜαʔϏεձһ਺ສਓಥഁ

  5.  ݴޠΧࠃ

  6.  ݴޠΧࠃ

  7. ϓϨϛΞϜձһ਺ ݄ؒར༻ऀ਺ʢࠃ಺ʣ ݄ؒར༻ऀ਺ʢւ֎ʣ Ϩγϐ਺ ສਓ ສ඼ ສ඼  ສਓ 

    ສਓ

  8. Ԋֵ ೥݄ ༗ݶձࣾίΠϯʢݱΫοΫύουגࣜձࣾʣઃཱ ೥݄ ೥݄ ೥݄ ೥݄ ೥݄ ೥݄ Ϩγϐͷ౤ߘɾݕࡧαʔϏεLJUDIFO!DPJO։࢝

    ΫοΫύουʹαʔϏε໊มߋ ೥݄ ϓϨϛΞϜαʔϏε։࢝ Ϛβʔζ্৔ ౦ূҰ෦ʹࢢ৔มߋ ϓϨϛΞϜαʔϏεձһ਺ສਓಥഁ ೥݄ ւ֎ల։Λຊ֨Խ ઃཱप೥ ೥݄ ݚڀ։ൃ෦໳৽ઃ

  9. എܠ ʕʕ·ͣ͸ɺݚڀ։ൃ෦͕ੜ·Εͨഎܠʹ͍ͭͯڭ͍͑ͯͩ͘͞ɻ ɹݚڀ։ൃ෦্ཱ͕͕͖͔͚ͪͬͨͬ͸ɺ೥͝Ζʹࣾ಺ͰػցֶशͳͲʹڵຯΛ࣋ͬͨࣾһ͕਺໊ू·ͬͯݚڀ ։ൃνʔϜΛ࡞Γɺۀ຿֎Ͱࣗओతʹٕज़ͷݚڀΛ͸͡Ίͨͱ͜Ζʹ͔͞ͷ΅Γ·͢ɻΫοΫύου͸αʔϏεͷӡӦ Λ௨ͯ͡ɺ͞·͟·ͳϢχʔΫͳσʔλΛอ༗͍ͯͨ͠ͷͰɺͦΕΒͱػցֶशΛ૊Έ߹ΘͤΔ͜ͱͰ໘ന͍αʔϏε ͕࡞Εͳ͍͔ͱߟ͍͑ͯͨͷͰ͢ɻ ɹ౰࣌ɺͦΕ͸ձࣾʹͱͬͯ͸͋͘·Ͱlඇެࣜzͳ׆ಈͩͬͨͷͰ͕͢ɺͨͱ͑͹ΫοΫύουͷϏοάσʔλΛݚڀ ໨తͱͯ͠େֶͷݚڀࣨʹఏڙ͢Δͱ͍ͬͨ׆ಈΛଓ͚ΔதͰɺࣾ಺Ͱ΋஫໨͞ΕΔΑ͏ʹͳΓ·ͨ͠ɻ͢Δͱɺ ೥͝Ζ͔Β͸ੈͷதͰػցֶशʹେ͖ͳ஫໨͕ू·ΔΑ͏ʹͳΓɺͦΕʹͱ΋ͳͬͯࣾ಺ͷ͞·͟·ͳ෦ॺ͔ ΒɺࢲͨͪʹػցֶशΛ׆༻ͨ͠αʔϏε΁ͷ։ൃࢧԉͷ੠ֻ͕͔ΔΑ͏ʹͳͬͨͷͰ͢ɻ

    ɹͦ͏ͳΔͱɺػցֶश͕ಘҙͳΤϯδχΞ͕ධՁ͞ΕΔ૊৫͕ࣾ಺ʹͳ͍ͱ͍͏՝୊͕ੜ·Εͯ͠·͍·͢ɻ౰࣌ɺ ͦ͏͍͏ΤϯδχΞ͸͞·͟·ͳ෦ॺʹϔϧϓʹ͍ͬͯ׆ಈ͍ͯͨ͠ͷͰɻͦ͏ͨ͠૊৫తͳ՝୊ͱɺੈͷதͷػց ֶशʹؔ͢ΔϜʔϒϝϯτͳͲΛഎܠʹɺ೥݄ʹݚڀ։ൃࣨͱͯ͠ਖ਼ࣜͳࣾ಺νʔϜ͕Ͱ͖ɺ݄ʹ෦ʹঢ֨ ͠·ͨ͠ɻ

  10. ؔ࿈αʔϏε ͓·͔ͤ੔ཧ ०ͷ໺ࡊϨγϐ΍ ϝΠϯʹ࢖͑ΔϨγϐ͕ ͙͢ʹݟ͔ͭΔ "MFYBεΩϧ ࢖͍͍ͨࡐྉ͚ͩͰ ͙͢ʹ࡞ΕΔਓؾͷϨγϐఏҊ ྉཧ͖Ζ͘ ࡱΔ͚ͩͰه࿥ʹʂ

    ͋ͳͨͷ͓ྉཧࣸਅΛ ࣗಈͰ੔ཧ Ϩγϐͷͦͷઌ΁ ͓͍͍͠εϚʔτΩονϯ 0J$Z

  11. ਓ਺ ʢׅހ಺͸.-ܥͷਓ਺ʣ        

         ਓ ਓ ਓ ਓ ਓ ਓ ϋʔυ΢ΣΞܥͷνʔϜ͕ ผ෦໳ͱͯ͠εϐϯΞ΢τ

  12. .-ܥͷτϐοΫʢҰྫʣ ࣗવݴޠॲཧ ɾ&ODPEFS%FDPEFSʹΑΔࡐྉ໊ͷਖ਼نԽ ɾ-45.ʹΑΔඇखॱͷݕग़ ɾGBTU5FYUʹΑΔࡐྉͷϨίϝϯυ ը૾ೝࣝ ɾ*ODFQUJPOWʹΑΔྉཧࣸਅͷநग़΍෼ྨ ɾ.PCJMF/FUʹΑΔݟӫ͑͢Δྉཧࣸਅͷநग़ ɾ43("/ʹΑΔྉཧࣸਅͷ௒ղ૾ .-0QT

    ɾ4MBDL$IBU#PUʹΑΔ&$Πϯελϯεͷ؅ཧ ɾ%PDLFSͱDPPLJFDVUUFSʹΑΔϓϩδΣΫτͷ؅ཧ ɾ&$4ʹΑΔΞϓϦͷσϓϩΠ΍%8)ʹΑΔσʔλͷ؅ཧ ΦʔϓϯαΠΤϯε ɾݚڀ༻Ϩγϐσʔλͷެ։ʢXࠃཱ৘ใֶݚڀॴʣ ɾ֤छίϯϖςΟγϣϯͷ࣮ࢪʢX಺ֳ෎ɺਓ޻஌ೳֶձɺʜʣ ɾ࿦จͷࣥචʢ೥Ͱຊ͕࠾࿥ʣ .-ܥ͸ਓ

  13. ৄ͘͠͸ BOENPSF

  14. ͝ਗ਼ௌ͋Γ͕ͱ͏͍͟͝·ͨ͠ɻ