Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Watson Studioで プログラミングなしの機械学習モデル作成

Kyoko Nishito
December 08, 2018

Watson Studioで プログラミングなしの機械学習モデル作成

2019年12月8日 IBM Cloud Community Summit 2018の講演資料です。

Kyoko Nishito

December 08, 2018
Tweet

More Decks by Kyoko Nishito

Other Decks in Technology

Transcript

  1. , / 4 Model Lifecycle Management dBytBny tyg r BotBny

    tyg r Bg hBv / L s v A Bny A o A Bg IP hBv • (at l WC • BoyfBehBvR (hB v pe S • ) •      • S BoyfBe dBy tBny tyg rA B otBny tyg r • cyi pBe ufBe • L M
  2.  5 dog cat cat dog Positive ID: Pug Unknown

    Unknown Positive ID: Pug dog dog dog dog  
  3. a M 6 dog cat cat dog Positive ID: Pug

    Unknown Unknown Positive ID: Pug  a M M S a L A W
  4. r vLW g e 7 lnN H W mcg e

    v LA u Pg e W ENLWg e o sG LG I g eP o LC DTS mc g e • ( , ,(-) GH t 3 • ( , GH 3 • ( , ,(-) t 3 • ( , 3 • , : i 3: : a n 3
  5. a 8 CDH E CDH ILH E ILH L GHA

    H N a • - ,( ) CDH E • - ) CDH • - ,( ) ILH E • - ) ILH • : ST : : : W P : : W
  6. I 10 ( b c ih • M ( •

    b B e b c k dgA k • / 1 )1 2 M ( • b B e b c k W A k • / 1 M 1.CI c • b B e b c k W A k • L S a S M a S 
  7. 21 2 M 12 C G ( ) 2 I

    B 21 2 B . 3 . M 
  8. ( 2 2 . 13 N D ) 42 )

    5 N 5 
  9. 17 ,0,1 14 _ l i ,0,1 1 14 c

    5 • 1 1 4 01, 30 4,0 ,., 7 o a ,0,1 1 14 5 • ,0,1 14 n c  a eg 
  10. A S 22 2 2 4 2 2 3 .

    3 4 AE I 1 
  11. M 23 55 .4 L A 44 + 5 4

    + 4 4 + 4 54 + . 
  12. S 24 81 61 L 81 61 .471 . 6

    . 1 . . 4 1 EM I M . . 6 L 
  13. 29 + 6 2 . 2 2 2 + 6

    2 . 2 2 2 + 
  14. ] d b 30 3. Na 56:3 3 :6:4 3

    6 3 b 56:3 3 :6:4 C[e M L Nc 383 38 A 3 38 68 3 383 : 6 3 b 53 7 N[e : 8N[e 3 3.N S 
  15. L 33 0 0 . b 61 b 61 a

    a F 61 C 61 
  16. l R c 34 i C • 2 (1 .

    6 32 f B M f • ) 16 1 (1 . 6 32 a f M f • 32 g i ef f C ) 16 1 (1 . 6 32 : 
  17. e a H 35 T H ) d H (.

    H H H T 6 1 
  18. i u Dn 38 e lg D B 6 LMRG

    cs 6 6 d m .6 6 Y I MT . 6 6 y MRG d m *. . tv Nar 6 . 6 o cs 6 6 F T cs
  19. mtlsg iPa c 54 o E3 [ [LS A A

    ADP P nN TSL H • ) - -). 1 I P • ) - ,(1 I Pr • )1 - -). 1 WP • )1 - ,(1 WPr • , A A A _cd • , A D A G [i • , A D : G]e 
  20. 60