Exploratory: ランダムフォレストの紹介

Slide 1

Slide 1 text

EXPLORATORY

Slide 2

Slide 2 text

2 εϐʔΧʔ ੢ా צҰ࿠ CEO EXPLORATORY ུྺ 2016೥ɺσʔλαΠΤϯεͷຽओԽͷͨΊɺExploratory, Inc Λཱͪ ্͛Δɻ Exploratory, Inc.ͰCEOΛ຿ΊΔ͔ͨΘΒɺσʔλαΠΤϯεɾϒʔ τΩϟϯϓɾτϨʔχϯάͳͲΛ௨ͯ͠γϦίϯόϨʔͰߦΘΕ͍ͯ Δ࠷ઌ୺ͷσʔλαΠΤϯεͷීٴͱڭҭʹऔΓ૊Ήɻ ถΦϥΫϧຊࣾͰɺ16೥ʹΘͨΓσʔλαΠΤϯεͷ։ൃνʔϜΛ཰ ͍ɺػցֶशɺϏοάɾσʔλɺϏδωεɾΠϯςϦδΣϯεɺσʔ λϕʔεʹؔ͢Δ਺ଟ͘ͷ੡඼ΛੈʹૹΓग़ͨ͠ɻ @KanAugust

Slide 3

Slide 3 text

Vision ΑΓΑ͍ҙࢥܾఆΛ͢ΔͨΊʹ σʔλΛ࢖͏͜ͱ͕౰ͨΓલʹͳΔ

Slide 4

Slide 4 text

Mission σʔλαΠΤϯεͷຽओԽ

Slide 5

Slide 5 text

5 ୈ̏ͷ೾ σʔλαΠΤϯεɺAIɺػցֶश͸౷ܭֶऀɺ։ൃऀͷͨΊ͚ͩͷ΋ͷͰ͸͋Γ·ͤΜɻ σʔλʹڵຯͷ͋ΔਓͳΒ୭΋͕ੈքͰ࠷ઌ୺ͷΞϧΰϦζϜΛ࢖ͬͯ ϏδωεσʔλΛ؆୯ʹ෼ੳͰ͖Δ΂͖Ͱ͢ɻ Exploratory͕ͦ͏ͨ͠ੈքΛՄೳʹ͠·͢ɻ

Slide 6

Slide 6 text

ୈ1ͷ೾ ୈ̎ͷ೾ ୈ̏ͷ೾ ϓϥΠϕʔτ(ߴ͍/ݹ͍) Φʔϓϯɾιʔε(ແྉ/࠷ઌ୺) UI & ϓϩάϥϛϯά ϓϩάϥϛϯά 2016 2000 1976 ϚωλΠθʔγϣϯ ίϞσΟςΟԽ ຽओԽ ౷ܭֶऀ σʔλαΠΤϯςΟετ Exploratory ΞϧΰϦζϜ Ϣʔβʔɾ ମݧ πʔϧ Φʔϓϯɾιʔε(ແྉ/࠷ઌ୺) UI & ࣗಈԽ ϏδωεɾϢʔβʔ ςʔϚ σʔλαΠΤϯεͷຽओԽ

Slide 7

Slide 7 text

質問 ExploratoryͰ؆୯ʹͰ͖ΔλεΫ 伝えるデータアクセスデータラングリング可視化アナリティクス統計/機械学習 UI

Slide 8

Slide 8 text

EXPLORATORY ΦϯϥΠϯɾηϛφʔ

Slide 9

Slide 9 text

Analytics ϥϯμϜϑΥϨετ

Slide 10

Slide 10 text

10 ͦͷલʹɻɻɻ

Slide 11

Slide 11 text

11 σʔλ෼ੳͱ͸ ૬ؔɺύλʔϯΛݟ͚ͭΔ͜ͱ

Slide 12

Slide 12 text

12 څྉ ೥ྸ ৬छ ۈଓ೥਺ ੑผ 10,000 60 Manager 24 Male 3,000 40 Sales Rep 3 Female 11,000 50 Research Director 35 Female 4,000 20 HR Rep 4 Male 5,000 30 HR Rep 5 Female 10,000 45 Manager 20 Female ஌Γ͍ͨ͜ͱ ଐੑσʔλ

Slide 13

Slide 13 text

13 ՄࢹԽʂ

Slide 14

Slide 14 text

څྉ vs. ৬छ

Slide 15

Slide 15 text

څྉ vs. ۈଓ೥਺

Slide 16

Slide 16 text

څྉ vs. ֊ڃ

Slide 17

Slide 17 text

17 σʔλ ૬ؔɾ ύλʔϯ ՄࢹԽͯ͠૬ؔɾύλʔϯΛҰͭҰͭ໨Ͱݟͯݕ஌͢Δ

Slide 18

Slide 18 text

18 ΊΜͲ͍͘͞ʂ

Slide 19

Slide 19 text

19 ΞφϦςΟΫεʂ

Slide 20

Slide 20 text

20 σʔλ ૬ؔɾ ύλʔϯ ػցֶशɾ౷ܭ ΞφϦςΟΫεΛ࢖ͬͯ૬ؔɾύλʔϯΛޮՌతʹݟ͚ͭΔɻ ΞφϦςΟΫε

Slide 21

Slide 21 text

21 ϥϯμϜ ϑΥϨετ Ϟσϧ ༧ଌϞσϧΛ࡞Δ σʔλ ΞϧΰϦζϜ

Slide 22

Slide 22 text

22 Monthly Income Age Job Role Department Gender ? 60 Manager Sales Male ? 40 Sales Rep R&D Female ? 30 Research Director HR Female Monthly Income Age Job Role Department Gender 10,000 60 Manager HR Male 11,000 40 Research Director R&D Female 4,000 30 HR Rep HR Female ༧ଌ͢Δ ౴͑ͷͳ͍σʔλ Ϟσϧ ϥϯμϜ ϑΥϨετ

Slide 23

Slide 23 text

23 Ϟσϧ͸σʔλͷதʹ͋ΔύλʔϯΛ΋ͱʹ࡞ΒΕΔ Ϟσϧ ϥϯμϜ ϑΥϨετ

Slide 24

Slide 24 text

24 Ͳͷม਺͕ΑΓ૬͕ؔ͋Δͷ͔ɺͲ͏͍͏ؔ܎ੑ Λ͍࣋ͬͯΔͷ͔Λ஌͍ͬͯΔɻ Ϟσϧ ϥϯμϜ ϑΥϨετ

Slide 25

Slide 25 text

25 σʔλ ΞφϦςΟΫεʹΑͬͯಘΒΕͨΠϯαΠτΛ ՄࢹԽ͢Δ͜ͱͰɺ௚ײతʹཧղ͢Δ ΞφϦςΟΫε ʢػցֶशɺ౷ܭʣ ૬ؔ / ύλʔϯ

Slide 26

Slide 26 text

26 ϥϯμϜϑΥϨετ

Slide 27

Slide 27 text

27 ϑΥϨετʢ৿ʣ

Slide 28

Slide 28 text

28 Ξϯαϯϒϧֶश • ෳ਺ͷϞσϧʹֶशͤ͞ɺͦΕͧΕͷ༧ଌ݁ՌΛҰ ͭͷ༧ଌʹ·ͱΊΔɻ • ྫɿRandom Forest, XGBoost

Slide 29

Slide 29 text

29 σʔλ ݁Ռ ܾఆ໦

Slide 30

Slide 30 text

30 ܾఆ໦ σʔλ αϯϓϧ αϯϓϧ αϯϓϧ ݁Ռ ݁Ռ ݁Ռ ଟ਺ܾ … ϥϯμϜαϯϓϧ ϥϯμϜϑΥϨετ

Slide 31

Slide 31 text

31 Mother Age Father Age Weight Plurality State Is Premature 40 42 5.5 1 CA TRUE 33 33 6.7 1 NY FALSE 32 36 7.0 1 WA FALSE 28 28 4.5 2 NC TRUE 24 26 6.0 1 MI FALSE 28 26 6.7 1 AZ FALSE 43 40 7.6 1 TX FALSE 38 33 4.2 2 FL TRUE 34 32 5.7 1 CA FALSE 29 33 5.2 1 NY TRUE ݩσʔλ

Slide 32

Slide 32 text

32 Mother Age Father Age Weight Plurality State is_premature 40 42 5.5 1 CA TRUE 33 33 6.7 1 NY FALSE 32 36 7.0 1 WA FALSE 28 28 4.5 2 NC TRUE 24 26 6.0 1 MI FALSE 28 26 6.7 1 AZ FALSE 43 40 7.6 1 TX FALSE 38 33 4.2 2 FL TRUE 34 32 5.7 1 CA FALSE 29 33 5.2 1 NY TRUE ༧ଌର৅ͷྻ: is_premature (ૣ࢈͔Ͳ͏͔)

Slide 33

Slide 33 text

33 Mother Age Weight is_premature 40 5.5 TRUE 33 6.7 FALSE 32 7.0 FALSE 28 4.5 TRUE σʔλ ߦͱྻΛϥϯμϜʹαϯϓϧ͢Δɻis_premature ྻ͸༧ଌର৅ͳͷͰͲͷαϯϓϧʹ΋ඞͣೖ Δɻ Mother Age Plurality State is_premature 28 1 AZ FALSE 43 1 TX FALSE 38 2 FL TRUE Father Age State is_premature 33 FL TRUE 32 CA FALSE 33 NY TRUE

Slide 34

Slide 34 text

34 Mother Age Weight is_premature 40 5.5 TRUE 33 6.7 FALSE 32 7.0 FALSE 28 4.5 TRUE Mother Age Plurality State is_premature 28 1 AZ FALSE 43 1 TX FALSE 38 2 FL TRUE Father Age State is_premature 33 FL TRUE 32 CA FALSE 33 NY TRUE σʔλ αϯϓϧ͞Εͨσʔλ͔ΒϞσϧΛ࡞੒͢Δɻ

Slide 35

Slide 35 text

35 σʔλ αϯϓϧ αϯϓϧ αϯϓϧ ݁Ռ ݁Ռ ݁Ռ ଟ਺ܾ …

Slide 36

Slide 36 text

36 ม਺ॏཁ౓ (Variable Importance)

Slide 37

Slide 37 text

37 • RandomForestͰ࡞ͬͨ༧ଌϞσϧ͔ΒɺͲͷม਺͕༧ଌͷࡍʹॏཁ ͳͷ͔ͱ͍͏৘ใ͕ͱΕΔɻ • ୳ࡧతσʔλ෼ੳʹΑ͘࢖ΘΕΔɻ

Slide 38

Slide 38 text

Let's Try!

Slide 39

Slide 39 text

39 ैۀһ͕ࣙΊΔͷʹԿ͕ؔ܎͍ͯ͠Δͷ͔ௐ΂͍ͨɻ

Slide 40

Slide 40 text

40 ΞφϦςΟΫεɾϏϡʔ

Slide 41

Slide 41 text

41 ༧ଌର৅ྻͷબ୒

Slide 42

Slide 42 text

42 ม਺ͷྻͷબ୒

Slide 43

Slide 43 text

43 શͯͷྻΛ༧ଌม਺ͱͯ͠બ୒

Slide 44

Slide 44 text

Slide 45

Slide 45 text

45 ༧ଌӨڹ౓

Slide 46

Slide 46 text

Slide 47

Slide 47 text

47 • ॏཁ౓্Ґͷม਺ʹ͍ͭͯɺPartial Dependencyͱݺ͹ ΕΔ৘ใΛՄࢹԽ͍ͯ͠Δɻ • ಛఆͷม਺ͷΈΛಈ͔͠ɺ࢒Γͷม਺Λݻఆͨ͠ͱ͖ʹɺ ݁ՌʹͲͷΑ͏ͳӨڹ͕͋Δ͔Λɺ࡞੒ͨ͠ϞσϧΛ΋ͱ ʹϓϩοτͨ͠΋ͷɻ ม਺ॏཁ౓ - ༧ଌӨڹ౓

Slide 48

Slide 48 text

48 ϞσϧͷαϚϦ

Slide 49

Slide 49 text

Slide 50

Slide 50 text

50 ༧ଌϚτϦοΫε

Slide 51

Slide 51 text

51 ༧ଌϚτϦοΫεϏϡʔ͸ ༧ଌ݁Ռͷਖ਼ղɺෆਖ਼ղΛ ύʔηϯτͰදࣔ͢Δɻ

Slide 52

Slide 52 text

52 Ϟσϧͷ༧ଌਫ਼౓Λ ධՁ͢ΔͨΊͷࢦඪ

Slide 53

Slide 53 text

53 Ϟσϧͷ༧ଌਫ਼౓ͱ͸ʁ ͲΕ͚࣮ͩࡍͷσʔλʹϚον͍ͯ͠Δ͔

Slide 54

Slide 54 text

54 ਖ਼ղ཰ʢAccuracy RateʣΛௐ΂Δ

Slide 55

Slide 55 text

55 TRUE FALSE TRUE 5 15 FALSE 15 195 ༧ଌͷ݁Ռ ࣮ࡍͷσʔλ

Slide 56

Slide 56 text

56 TRUE FALSE TRUE 5 15 FALSE 15 195 ࣮ࡍͷ݁Ռ ਖ਼ղ཰ = (5 + 195) / 240 = 0.875 ༧ଌͷ݁Ռ

Slide 57

Slide 57 text

57 ਖ਼ղ཰͑͞Α͚Ε͹ͦΕͰ͍͍ͷ͔ʁ

Slide 58

Slide 58 text

58 TRUE FALSE TRUE True Positive λΠϓ2 Τϥʔ (ୈೋछաޡ) FALSE λΠϓ1 Τϥʔ (ୈҰछաޡ) True Negative ࣮ࡍͷ஋ ༧ଌͷ݁Ռ

Slide 59

Slide 59 text

59 ͋ͳͨ೛৷ ͯ͠·͢Ͷ ͋ͳͨ೛৷ ͯ͠·ͤΜΑ

Slide 60

Slide 60 text

60 TRUE FALSE TRUE 5 15 FALSE 15 200 ࣮ࡍͷ஋ Precisionʢద߹཰ʣ Precision = 5 / (5+15) = 25% ༧ଌͷ݁Ռ

Slide 61

Slide 61 text

61 TRUE FALSE TRUE 5 15 FALSE 15 200 ࣮ࡍͷ஋ TRUEͱ༧ଌ͞Ε͕ͨɺͦΕ͕ਖ਼͍͠ͷ͸25%͚ͩͩͬͨɻ (5 / (5+15)). Precision = 5 / (5+15) = 25% ༧ଌͷ݁Ռ

Slide 62

Slide 62 text

62 TRUE FALSE TRUE 5 15 FALSE 15 200 ࣮ࡍͷ஋ TRUEͱ༧ଌ͞Ε͕ͨɺͦΕ͕ਖ਼͍͠ͷ͸25%͚ͩͩͬͨɻ (5 / (5+15)). Precision = 5 / (5+15) = 25% ༧ଌͷ݁Ռ λΠϓ̍Τϥʔ͕75%΋ى͖͍ͯΔ

Slide 63

Slide 63 text

63 ΋͠TRUEͱ༧ଌ͢ΔͳΒɺ݁Ռ͕ຊ౰ʹ TRUEͰ͋Δ΂͖ɻ λΠϓ1 Τϥʔ (ୈҰछաޡ) ͋ͳͨ೛৷ ͯ͠·͢Ͷ

Slide 64

Slide 64 text

64 Hi Grandma Kan΁ ࠓͲ͜ʹ͍·͔͢ʁେৎ෉Ͱ͔͢ʁ௕͍ؒ͋ͳ͔ͨΒͷฦ৴͕ͳ͍ ͷͰ৺഑Ͱ͢ɻԿ΋ͳ͍ͱ͍͍Μ͚ͩͲɻޙͰɺ΋͏Ұ౓࿈བྷͯ͘͠ ͍ͩ͞ɻ ͓͹͋ͪΌΜΑΓ εύϜ ΋͠TRUEͱ༧ଌ͢ΔͳΒɺ  ݁Ռ͕ຊ౰ʹTRUEͰ͋Δ΂͖ɻ λΠϓ1 Τϥʔ (ୈҰछաޡ)

Slide 65

Slide 65 text

65 TRUE FALSE TRUE 5 15 FALSE 15 200 ࣮ࡍͷ஋ Recallʢݕग़཰ʣ Recall = 5 / (5+15) = 25% ༧ଌͷ݁Ռ

Slide 66

Slide 66 text

66 TRUE FALSE TRUE 5 15 FALSE 15 200 ࣮ࡍͷ஋ ࣮ࡍʹTRUEͷͱ͖ʹɺͦΕΛ༧ଌग़དྷͨͷ͸25%͚ͩͩͬͨɻ Recall = 5 / (5+15) = 25% ༧ଌͷ݁Ռ

Slide 67

Slide 67 text

67 TRUE FALSE TRUE 5 15 FALSE 15 200 ࣮ࡍͷ஋ ࣮ࡍʹTRUEͷͱ͖ʹɺͦΕΛ༧ଌग़དྷͨͷ͸25%͚ͩͩͬͨɻ Recall = 5 / (5+15) = 25% ༧ଌͷ݁Ռ λΠϓ̎Τϥʔ͕75%΋ى͖͍ͯΔ

Slide 68

Slide 68 text

68 FALSEͱ༧ଌ͢ΔΑΓ͸TRUEͱ༧ଌ ͢Δํ͕·͠ɻ TRUEͷ࣌ʹɺͦΕΛTRUEͱ༧ଌ͠ ͳ͔ͬͨ৔߹ɺॏେͳଛ֐ΛҾ͖ى͜ ͢Մೳੑ͕͋Δɻ λΠϓ2 Τϥʔ (ୈೋछաޡ) ͋ͳͨ೛৷ ͯ͠·ͤΜ

Slide 69

Slide 69 text

69 F Score: 0͔Β1ͷؒͷ஋ΛͱΓɺ1ʹ͍ۙ΄Ͳྑ͍ɻ F Score RecallͱPrecisionͷฏۉʢHarmonic Mean) 0.25

Slide 70

Slide 70 text

70 ภͬͨ༧ଌ݁Ռ ༧ଌϚτϦοΫεΛݟΔͱɺϞσϧ͸ ୯ʹ΄ͱΜͲ͍ͭͰ΋FALSEͱ༧ଌͯ͠ ͍Δ͜ͱ͕෼͔Δɻ

Slide 71

Slide 71 text

71 ෆۉߧͳσʔλ ࣙΊͨैۀһͷׂ߹͸ѹ౗తʹগͳ͍ɻ

Slide 72

Slide 72 text

72 ෆۉߧͳσʔλʹΑΔ༧ଌͷ໰୊ • FALSEͷσʔλ͕ଟ͍ͨΊɺͦͷΑ͏ͳ ଟ਺೿ͷ஋Λ౴͑ͱ͢Δ͜ͱͰਖ਼ղ͢Δ ֬཰Λ࠷େʹ͍ͯ͠ΔΑ͏ͩɻ • λΠϓ2 Τϥʔ͕ଟ͍Ϟσϧʹͳͬͯ͠ ·͍ͬͯΔɻ

Slide 73

Slide 73 text

73 ͜ΕΒͷ໰୊͸ɺαϚϦɾϏϡʔͷࢦඪʹ΋ݱΕΔɻ Recall ͸࣮ࡍTRUEͷ৔߹ʹɺTRUEͩͱ༧ଌग़དྷׂͨ߹ɻ  ͜ͷࢦඪ͕௿͍ͱ͖͸ɺλΠϓ2 Τϥʔ͕ଟ͘ى͖Δ͜ͱ Λࣔ͢ɻ

Slide 74

Slide 74 text

74 σʔλͷෆۉߧΛղফ͢Δ

Slide 75

Slide 75 text

75 ଟ਺೿σʔλΛαϯϓϦϯάͯ͠ݮΒ͠ʢΞϯμʔɾα ϯϓϦϯάʣɺগ਺೿σʔλΛ߹੒ͯ͠૿΍͢ʢΦʔ όʔɾαϯϓϦϯάʣɻ ΦʔόʔɾαϯϓϦϯά ΞϯμʔɾαϯϓϦϯά গ਺೿σʔλΛ ߹੒ͯ͠૿΍͢ গ਺೿σʔλΛ αϯϓϦϯάͯ͠ ݮΒ͢ ݩσʔλ ݩσʔλ

Slide 76

Slide 76 text

76 • গ਺೿σʔλΛ߹੒ͯ͠૿΍͢ʢΦʔόʔɾαϯϓϦϯάʣͷ ͨΊʹɺSMOTE (Synthetic Minority Oversampling Technique) ͱ ͍͏ΞϧΰϦζϜ͕࢖͑Δɻ • ExploratoryͰ͸ɺΞφϦςΟΫεɾϏϡʔͷதͷϓϩύςΟ͔ Βઃఆ͢Δ͜ͱ͕Ͱ͖Δɻ • ·ͨɺσʔλϥϯάϦϯάͷεςοϓͱͯ͠ߦ͏͜ͱ΋Ͱ͖ Δɻ SMOTEʹΑΔগ਺೿σʔλͷ߹੒

Slide 77

Slide 77 text

77 SMOTE • গ਺೿σʔλΛϥϯμϜʹબͿ • બ͹Εͨগ਺೿σʔλʹ࠷΋͍ۙɺଞͷগ਺೿σʔ λΛݟ͚ͭΔɻ • 2ͭͷσʔλͷதؒ఺ΛϥϯμϜʹબͼɺ2ͭͷσʔ λ͔Βͷ஋Λൺྫ഑෼ͨ͠஋ΛׂΓ౰ͯΔɻ

Slide 78

Slide 78 text

YesΛબΜͰσʔλͷෆۉߧΛղফ ͢ΔͨΊͷલॲཧΛ͢Δɻ

Slide 79

Slide 79 text

• Ϟσϧ͕TRUEͱ༧ଌ͢Δׂ߹͕૿ ͍͑ͯΔɻ • ౴͕͑TRUEͰ͋Δ৔߹Ͱ΋͔ͬ͠ Γͱ༧ଌͰ͖͍ͯΔɻ

Slide 80

Slide 80 text

RecallͱɺRecallͱPrecisionͷฏۉͰ͋Δ F Score͕վળ͞Ε͍ͯΔɻ

Slide 81

Slide 81 text

ઌ΄Ͳͱ͸ɺएׯɺॱҐ͕มΘ͍ͬͯΔɻ

Slide 82

Slide 82 text

TRUEͷ֬཰্͕͕͍ͬͯΔɻ

Slide 83

Slide 83 text

83 ෆۉߧΛௐ੔͢Δલ

Slide 84

Slide 84 text

84 ม਺ॏཁ౓ͷ໰୊

Slide 85

Slide 85 text

• ݁ہɺͲͷม਺͕Өڹ͋Δͷ͔͕Θ͔Βͳ͍ɻ • ϥϯμϜϑΥϨετͷʮϥϯμϜੑʯʹΑͬͯɺྲྀͨ͢ͼʹม਺ॏཁ ౓ͷॱҐ͕ҧͬͯ͘Δɻ 85

Slide 86

Slide 86 text

86 Ͳͷม਺͕ॏཁͳͷ͔?

Slide 87

Slide 87 text

87 ܾఆ໦ σʔλ αϯϓϧ αϯϓϧ αϯϓϧ ౤ථ ౤ථ ౤ථ ݁࿦ … ϥϯμϜαϯϓϧ ϥϯμϜαϯϓϧΛར༻͍ͯ͠ΔͷͰɺ݁ՌʹϥϯμϜੑ͕͋Δɻ

Slide 88

Slide 88 text

88 • ݁ՌʹϥϯμϜੑ͕͋ΔͷͳΒɺॏཁͩͱ͞Εͨม਺΋ۮ વࠓճͦ͏͍͏݁Ռʹͳ͚ͬͨͩͰ͸ʁ • ม਺ͷॏཁ౓͕ۮવͳͷ͔ɺ౷ܭతʹݕఆͯ͠ΈΕ͹Α ͍ɻ Ϙϧʔλ

Slide 89

Slide 89 text

89 ϘϧʔλΛ༗ޮԽ

Slide 90

Slide 90 text

Slide 91

Slide 91 text

91 20ճϥϯμϜϑΥϨετͷϞσϧΛ࡞ͬͨͱ ͖ͷ݁Ռͷ஋ͷ෼෍Λശͻ͛ਤͰද͢ɻ

Slide 92

Slide 92 text

92 ͦΕͧΕͷม਺͕༧ଌͷ໾ʹͨͭͷ͔Ͳ͏ ͔ͷ౷ܭతͳ൑அΛ৭Ͱද͢ɻ

Slide 93

Slide 93 text

93 ༧ଌͷ໾ʹཱͭͱ౷ܭత ʹ൑அ͞Εͨม਺

Slide 94

Slide 94 text

94 ༧ଌͷ໾ʹͨͭͷ͔Ͳ͏͔  ·ͩ൑அग़དྷ͍ͯͳ͍ม਺

Slide 95

Slide 95 text

95 ༧ଌͷ໾ʹཱͨͳ͍ͱ౷ܭత ʹ൑அ͞Εͨม਺

Slide 96

Slide 96 text

૬ؔͷՄࢹԽ 96

Slide 97

Slide 97 text

97 σʔλ ΞφϦςΟΫεʹΑͬͯಘΒΕͨΠϯαΠτΛ ՄࢹԽ͢Δ͜ͱͰɺ௚ײతʹཧղ͢Δ ΞφϦςΟΫε ʢػցֶशɺ౷ܭʣ ૬ؔ / ύλʔϯ

Slide 98

Slide 98 text

98 νϟʔτΛ࢖͔ͬͯ֬ΊΔ

Slide 99

Slide 99 text

99 ͲͷνϟʔτΛ࢖͑͹Α͍ͷ͔ʁ

Slide 100

Slide 100 text

100 σʔλͷλΠϓͷ૊Έ߹Θͤ࣍ୈʂ

Slide 101

Slide 101 text

101 • ΧςΰϦʔ vs. ਺஋ • ਺஋ vs. ਺஋ • ΧςΰϦʔ vs. ΧςΰϦʔ σʔλͷλΠϓͷ૊Έ߹Θͤ

Slide 102

Slide 102 text

102 • ΧςΰϦʔ vs. ΧςΰϦʔ • ΧςΰϦʔ vs. ਺஋ • ਺஋ vs. ਺஋ ૬ؔΛՄࢹԽ͢Δ

Slide 103

Slide 103 text

103 ΧςΰϦʔ vs. ΧςΰϦʔ ΧςΰϦʔಉ࢜ͷؔ܎ΛݟΔʹ͸ɺΧςΰϦʔͷ ૊Έ߹Θͤ͝ͱʹ਺Λ਺্͑͛Δ͔ɺͦͷશମʹ ର͢Δׂ߹Λܭࢉͦ͠ͷ݁ՌΛՄࢹԽ͢Δɻ

Slide 104

Slide 104 text

104 ώʔτϚοϓ όϒϧ ελοΫɾόʔ

Slide 105

Slide 105 text

Overtime (Category) vs. Attrition (Category/Logical)

Slide 106

Slide 106 text

Job Level (Category) vs. Attrition (Category/Logical)

Slide 107

Slide 107 text

107 • ΧςΰϦʔ vs. ΧςΰϦʔ • ΧςΰϦʔ vs. ਺஋ • ਺஋ vs. ਺஋ ૬ؔΛՄࢹԽ͢Δ

Slide 108

Slide 108 text

108 ശͻ͛ਤ όΠΦϦϯਤ ີ౓ۂઢ

Slide 109

Slide 109 text

Monthly Income (Numeric) vs. Attrition (Category/Logical)

Slide 110

Slide 110 text

Monthly Income (Numeric) vs. Attrition (Category/Logical)

Slide 111

Slide 111 text

111 • ΧςΰϦʔ vs. ΧςΰϦʔ • ΧςΰϦʔ vs. ਺஋ • ਺஋ vs. ਺஋ ૬ؔΛՄࢹԽ͢Δ

Slide 112

Slide 112 text

112 ࢄ෍ਤ

Slide 113

Slide 113 text

No content

Slide 114

Slide 114 text

• ϓϩάϥϛϯάͳ͠ RݴޠͷUIͰ͋ΔExploratoryΛ෼ੳπʔϧͱͯ͠࢖༻͢ΔͨΊडߨத͸ɺϏδωεͷ ໰୊Λղܾ͢ΔͨΊʹඞཁͳσʔλαΠΤϯεͷख๏ͷशಘʹ100ˋूதͰ͖Δ • ෼ੳπʔϧͷϕϯμʔϩοΫΠϯͳ͠ ExploratoryͰͷ࡞ۀ͸શͯಠཱͨ͠ΦʔϓϯιʔεͷR؀ڥͰ࠶ݱ͕Մೳ • ࢥߟྗͱεΩϧͷशಘ σʔλαΠΤϯεͷεΩϧशಘ͚ͩͰͳ͘ɺσʔλ෼ੳʹඞཁͳࢥߟྗ΋शಘͰ͖Δ ಛ௃

Slide 115

Slide 115 text

Q & A

Slide 116

Slide 116 text

࿈བྷઌ ϝʔϧ [email protected] ΢ΣϒαΠτ https://ja.exploratory.io ϒʔτΩϟϯϓɾτϨʔχϯά https://ja.exploratory.io/training-jp Twitter @KanAugust