学習・推論パイプラインを構築する上で大切にしていること

by Takanobu Nozawa

Slide 1

Slide 1 text

ֶशɾਪ࿦ύΠϓϥΠϯΛߏங͢Δ্Ͱ େ੾ʹ͍ͯ͠Δ͜ͱ $POOFIJUP*OD໺ᖒ఩র ෼ੳίϯϖ-5ձ

Slide 2

Slide 2 text

͜Μʹͪ͸ʂ

Slide 3

Slide 3 text

ࠓ೔ൃද͢Δ͜ͱ͸Լهͱएׯॏෳ͠·͢ IUUQTTQFBLFSEFDLDPNUBLBQZEFUBGFOYJLPOQFOJPJUFUF[IFOHMJBOH HVBOMJOJQJCJTJUFJSVRVBOSFOMFJOJDIVBOFUBJYJBOHJ

Slide 4

Slide 4 text

ΞδΣϯμ ࣗݾ঺հ ύΠϓϥΠϯ͕ͳ͔ͬͨ࣌ͷπϥϛ େ੾ʹ͍ͯ͠Δ͜ͱ ‣ ֶशͷ࠶ݱੑ ‣ 1%$"ͷߴ଎Խ ‣ ҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έ ࣮૷ৄࡉ

Slide 5

Slide 5 text

ࣗݾ঺հ

Slide 6

Slide 6 text

ࣗݾ঺հ ໊લɿ໺ᖒ఩রʢ/P[BXB5BLBOPCVʣ ॴଐɿίωώτגࣜձࣾ ɹɹɿ͔ͨͺ͍!UBLBQZ w ʙίωώτʹ.-ΤϯδχΞͱͯ͠+0*/ w ػցֶशʢ/-1ɺਪનγεςϜʣΛϝΠϯʹ΍ΓͭͭΠϯϑϥʢ"84ʣ΋ษڧத w ,BHHMFͨ͠ΓɺϒϩάʢIUUQTXXXUBLBQZXPSLʣॻ͍ͨΓɺ໺ٿͨ͠Γɺ ϥʔϝϯ৯΂ͨΓ͍ͯ͠·͢ w ΦεεϝͷυϥϜࣜચ୕ػ͋ͬͨΒڭ͍͑ͯͩ͘͞

Slide 7

Slide 7 text

ࣗݾ঺հ ˞ʮӾཡ਺ʯʮར༻ऀ਺ʯ͸ϝσΟΞͱΞϓϦͷ߹ܭ஋ʢ೥݄݄ͷฏۉ஋ʣ ˞ʮϚϚ޲͚/PΞϓϦʯ͸೥݄Πϯςʔδௐ΂ɹௐࠪର৅ɿ೛৷தʙ̎ࡀ̌ϲ݄ͷࢠڙΛ࣋ͭঁੑ O Λநग़ ˞*OTUBHSBNͷϑΥϩϫʔ਺ɺ'BDFCPPLͷ͍͍Ͷ਺ɺ-*/&ͷͱ΋ͩͪ਺ͷ߹ܭ஋ ೥݄࣌఺ ϚϚϦ ΞϓϦɾ8FC 4/4 *OTUBHSBNɾ-*/&ɾ'BDFCPPL هࣄ ϚϚಉ࢜Ͱ೰ΈΛ૬ஊ͠߹͏2"ίϛϡχςΟΛத৺ʹ ϢʔβʔΛ֦େ͍ͯ͠·͢ ʮϚϚϦʯͰϢʔβʔಉ͕࢜ ͲΜͲΜܨ͕͍ͬͯ·͢ ϚϚͷੜ׆ʹ໾ཱͭهࣄΛ ෯޿͍δϟϯϧͰ഑৴͍ͯ͠·͢ ϚϚ޲͚/P̍ΞϓϦʹબग़ ਓͷϚϚ͕બͿʮݱࡏ࢖͍ͬͯΔΞϓϦʯʹ ͯɺ߲໨ ଞͷϚϚʹΦεεϝ͍ͨ͠ɺೝ஌౓ɺ ར༻཰ɺརศੑɺ޷ײ౓ Ͱ̍ҐΛ֫ಘ͠·ͨ͠ هࣄ਺ 6,000 هࣄҎ্ ྦྷܭϑΝϯ਺ ໿ 85 ສਓ ˞ ݄ؒӾཡ਺ ໿ 1.5ԯճ ˞ ݄ؒར༻ऀ਺ ໿ 650ສਓ ˞ ˞ l೰ΈzͱzڞײzΛ࣠ʹϚϚʹدΓఴ͍ ΞϓϦɾ8FCɾ4/4ͱଟ֯తʹαʔϏεΛల։͍ͯ͠·͢

Slide 8

Slide 8 text

ࣗݾ঺հ 0 450,000 900,000 1,350,000 1,800,000 2014/4 2014/5 2014/6 2014/7 2014/8 2014/9 2014/10 2014/11 2014/12 2015/1 2015/2 2015/3 2015/4 2015/5 2015/6 2015/7 2015/8 2015/9 2015/10 2015/11 2015/12 2016/1 2016/2 2016/3 2016/4 2016/5 2016/6 2016/7 2016/8 2016/9 2016/10 2016/11 2016/12 2017/1 2017/2 2017/3 2017/4 2017/5 2017/6 2017/7 2017/8 2017/9 2017/10 2017/11 2017/12 2018/1 2018/2 2018/3 2018/4 2018/5 2018/6 2018/7 2018/8 ݄ؒ౤ߘ਺ ໿ 150ສ݅ िʹ೔Ҏ্ىಈ͢Δ ΞΫςΟϒϢʔβʔ ໿ 50 ਓʹਓ 57$. ์ө ΞϓϦ૯%-਺ສ ਓʹਓ ਓʹਓ ਓʹਓ ਓʹਓ ˞ ˞ʮϚϚϦʯ಺ͷग़࢈༧ఆ೔Λઃఆͨ͠Ϣʔβʔ਺ͱɺްੜ࿑ಇলൃදʮਓޱಈଶ౷ܭʯͷग़ੜ਺͔Βࢉग़ ˞िʹճҎ্ىಈ͢ΔϢʔβʔ ˞ ೥ʹग़࢈ͨ͠ϚϚͷʮਓʹਓʯ͕ϚϚϦΛར༻த ೔ຊ࠷େڃن໛ΛތΔϒϥϯυ΁ͱ੒௕͍ͯ͠·͢ ˞

Slide 9

Slide 9 text

ύΠϓϥΠϯ͕ͳ͔ͬͨ࣌ͷπϥϛ

Slide 10

Slide 10 text

ύΠϓϥΠϯ͕ͳ͔ͬͨ࣌ͷπϥϛ ΊͬͪΌྑ͍είΞͷOPUFCPPL͕׬੒ ˣ ͜ͷOPUFCPPLΛ%VQMJDBUFͯ͠ɺ΋ͬͱྑ͍Ϟσϧ࡞ͬͪΌ͏ͧʂ

Slide 11

Slide 11 text

ҰํɺOPUFCPPLͷத਎͸ʜ

Slide 12

Slide 12 text

ύΠϓϥΠϯ͕ͳ͔ͬͨ࣌ͷπϥϛ <> import numpy as np import pandas as pd OPUFCPPLͷத਎ ɾ ɾ ɾ <> hogehoge <> hogehoge <> hogehoge ɾ ɾ ɾ

Slide 13

Slide 13 text

ύΠϓϥΠϯ͕ͳ͔ͬͨ࣌ͷπϥϛ <> import numpy as np import pandas as pd OPUFCPPLͷத਎ ɾ ɾ ɾ <> hogehoge <> hogehoge <> hogehoge ɾ ɾ ɾ ηϧͷ࣮ߦॱ͕ͪ͝Όͪ͝Ό ˣ ࠶ݱੑ͕ͳ͍

Slide 14

Slide 14 text

ύΠϓϥΠϯ͕ͳ͔ͬͨ࣌ͷπϥϛ <> import numpy as np import pandas as pd OPUFCPPLͷத਎ ɾ ɾ ɾ <> submission.to_csv('submission.csv', index=False)

Slide 15

Slide 15 text

ύΠϓϥΠϯ͕ͳ͔ͬͨ࣌ͷπϥϛ <> import numpy as np import pandas as pd OPUFCPPLͷத਎ ɾ ɾ ɾ <> submission.to_csv('submission.csv', index=False) ηϧ͕ଟ͘ɺಉ͡ܭࢉΛෳ਺࣮ߦ͠ͳ͚Ε͹ͳΒͳ͍ ˣ 1%$"͕஗͘ͳΓ͕ͪ

Slide 16

Slide 16 text

ύΠϓϥΠϯ͕ͳ͔ͬͨ࣌ͷπϥϛ dOPUFCPPL-JHIU#(.@TDPSF@$PQZJQZOC dOPUFCPPL-JHIU#(.@TDPSF@$PQZJQZOC dOPUFCPPL-JHIU#(.@TDPSF@$PQZJQZOC dOPUFCPPL-JHIU#(.@TDPSF@$PQZJQZOC dOPUFCPPL-JHIU#(.@TDPSF@$PQZJQZOC dOPUFCPPL-JHIU#(.@TDPSF@$PQZJQZOC dOPUFCPPL-JHIU#(.@TDPSF@$PQZJQZOC dOPUFCPPL-JHIU#(.@TDPSF@$PQZJQZOC dOPUFCPPL-JHIU#(.@TDPSF@JQZOC dOPUFCPPL-JHIU#(.@TDPSF@JQZOC dOPUFCPPL-JHIU#(.@TDPSF@JQZOC dOPUFCPPL-JHIU#(.@TDPSF@JQZOC dOPUFCPPL-JHIU#(.@TDPSF@JQZOC dOPUFCPPL-JHIU#(.@TDPSF@JQZOC dOPUFCPPL-JHIU#(.@TDPSF@JQZOC dOPUFCPPL-JHIU#(.@TDPSF@JQZOC ʜʜʜʜʜ

Slide 17

Slide 17 text

Slide 18

Slide 18 text

͜ΕΒͷπϥϛΛղফ͢΂͘ ୤ɾOPUFCPPLʹνϟϨϯδ ˣ ͦΜͳதͰେ੾ʹ͍ͯ͠Δ͜ͱΛ͓఻͑͠·͢ ˞&%"ͳͲ͸OPUFCPPLΛ࢖༻

Slide 19

Slide 19 text

౔୆͸͜ͷຊͰ͢

Slide 20

Slide 20 text

େ੾ʹ͍ͯ͠Δ͜ͱ

Slide 21

Slide 21 text

ͦͷ ʙֶशͷ࠶ݱੑʙ

Slide 22

Slide 22 text

ͦͷ ʙֶशͷ࠶ݱੑʙ ϙΠϯτ͸̏ͭ

Slide 23

Slide 23 text

େ੾ʹ͍ͯ͠Δ͜ͱ w ͲΜͳಛ௃ྔΛ࢖ͬͯ ʙֶशͷ࠶ݱੑʙ

Slide 24

Slide 24 text

େ੾ʹ͍ͯ͠Δ͜ͱ w ͲΜͳಛ௃ྔΛ࢖ͬͯ w ͲΜͳύϥϝʔλʔΛ࢖ͬͯ ʙֶशͷ࠶ݱੑʙ

Slide 25

Slide 25 text

େ੾ʹ͍ͯ͠Δ͜ͱ w ͲΜͳಛ௃ྔΛ࢖ͬͯ w ͲΜͳύϥϝʔλʔΛ࢖ͬͯ w ͲΜͳ$7Λ࢖ͬͯ ֶश͔ͨ͠Λอଘ͓ͯ͘͠ɻ ʙֶशͷ࠶ݱੑʙ

Slide 26

Slide 26 text

େ੾ʹ͍ͯ͠Δ͜ͱ w ͲΜͳಛ௃ྔΛ࢖ͬͯ w ͲΜͳύϥϝʔλʔΛ࢖ͬͯ w ͲΜͳ$7Λ࢖ͬͯ ֶश͔ͨ͠Λอଘ͓ͯ͘͠ɻ ʙֶशͷ࠶ݱੑʙ %0//"อଘଇ

Slide 27

Slide 27 text

ͦͷ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ

Slide 28

Slide 28 text

େ੾ʹ͍ͯ͠Δ͜ͱ w ͭͷQZUIPOεΫϦϓτͰ%0//"ΛঠѲ ࢖༻͢Δಛ௃ྔ ࢖༻͢Δύϥϝʔλ ࢖༻͢Δ$7 ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ

Slide 29

Slide 29 text

େ੾ʹ͍ͯ͠Δ͜ͱ w ͭͷQZUIPOεΫϦϓτͰ%0//"ΛঠѲ ࢖༻͢Δಛ௃ྔ ࢖༻͢Δύϥϝʔλ ࢖༻͢Δ$7 w GFBUVSFJNQPSUBODFͷݟ͑ΔԽ ࣍ͷ࣮ݧ΁ͷצॴΛ͔ͭΊΔΑ͏ʹ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ

Slide 30

Slide 30 text

ͦͷ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ

Slide 31

Slide 31 text

େ੾ʹ͍ͯ͠Δ͜ͱ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ w %0///"Ϟσϧ΍ϩάϑΝΠϧΛɺ࣮ߦ͝ͱʹʮ೔࣌ʯ TV⒏YΛ͚ͭͨσΟϨΫτϦɾϑΝΠϧͰࣗಈ؅ཧ ࢖༻͢Δಛ௃ྔ ύϥϝʔλ $7 ֶशͨ͠Ϟσϧ ֶशϩάϑΝΠϧ GFBUVSFJNQPSUBODF

Slide 32

Slide 32 text

େ੾ʹ͍ͯ͠Δ͜ͱ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ w %0///"Ϟσϧ΍ϩάϑΝΠϧΛɺ࣮ߦ͝ͱʹʮ೔࣌ʯ TV⒏YΛ͚ͭͨσΟϨΫτϦɾϑΝΠϧͰࣗಈ؅ཧ ࢖༻͢Δಛ௃ྔ ύϥϝʔλ $7 ֶशͨ͠Ϟσϧ ֶशϩάϑΝΠϧ GFBUVSFJNQPSUBODF w 1VCMJDείΞΛσΟϨΫτϦͷ1SFpYʹ෇͚Δ͜ͱͰɺ 1VCMJDͱMPDBMͷείΞͷରԠΛ෼͔Γ΍͘͢ʢৄࡉ͸ޙ΄Ͳʣ

Slide 33

Slide 33 text

Slide 34

Slide 34 text

࣮૷ৄࡉ

Slide 35

Slide 35 text

ྫʣIPHFQZ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'KFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, } runner = Runner(run_name, ModelLGB, features, dataset.get('target'), params_lgb, cv, FEATURE_DIR_NAME, MODEL_DIR_NAME) runner.run_train_cv() # ֶश runner.run_predict_cv() # ਪ࿦ Submission.create_submission(run_name) # submit࡞੒

Slide 36

Slide 36 text

Slide 37

Slide 37 text

Slide 38

Slide 38 text

Slide 39

Slide 39 text

Slide 40

Slide 40 text

Slide 41

Slide 41 text

Slide 42

Slide 42 text

ʙֶशͷ࠶ݱੑΛ୲อ͢ΔͨΊͷ޻෉ʙ

Slide 43

Slide 43 text

Slide 44

Slide 44 text

Slide 45

Slide 45 text

࣮૷ৄࡉ IPHFQZ ʙֶशͷ࠶ݱੑʙ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'KFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, } { "use_features": [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ], "model_params": { "boosting_type": "gbdt", "objective": "fair", "metric": "fair", "num_boost_round": 20000, "early_stopping_rounds": 1000, "verbose": 1000, "random_state": 999 }, "cv": { "method": "KFold", "n_splits": 5, "random_state": 42, "shuffle": true }, "dataset": { "run_name": "lgb_1128_2003", "feature_directory": "../data/features/remove_outlier/", "target": "salary" } } IPHF@QBSBNKTPO QZUIPOIPHFQZ

Slide 46

Slide 46 text

Slide 47

Slide 47 text

Slide 48

Slide 48 text

ʙߴ଎ͳ1%$"Λճͨ͢Ίͷ޻෉ʙ

Slide 49

Slide 49 text

Slide 50

Slide 50 text

Slide 51

Slide 51 text

࣮૷ৄࡉ IPHFQZ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'KFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, } ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ features = [ "age", "pclass", # "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'KFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, } IPHFQZ ର৅ͷಛ௃ྔΛίϝϯτΞ΢τ

Slide 52

Slide 52 text

Slide 53

Slide 53 text

Slide 54

Slide 54 text

Slide 55

Slide 55 text

Slide 56

Slide 56 text

࣮૷ৄࡉ IPHFQZ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'KFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, } ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'GroupKFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, 'cv_target':'user' } IPHFQZ DWͷهࡌΛมߋ

Slide 57

Slide 57 text

࣮૷ৄࡉ IPHFQZ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'KFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, } ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'GroupKFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, 'cv_target':'user' } IPHFQZ DWͷهࡌΛมߋ QZUIPOIPHFQZ

Slide 58

Slide 58 text

࣮૷ৄࡉ IPHFQZ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'KFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, } ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'GroupKFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, 'cv_target':'user' } IPHFQZ DWͷهࡌΛมߋ QZUIPOIPHFQZ ؆୯

Slide 59

Slide 59 text

࣮૷ৄࡉ IPHFQZ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'KFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, } ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ features = [ "age", "pclass", "family_size", "fare", "sibsp", "parch", "cabin" ] params_lgb = { 'boosting_type': 'gbdt', 'objective': 'fair', 'metric': 'fair', 'num_boost_round': 20000, 'early_stopping_rounds': 1000, 'verbose': 1000, 'random_state': 999 } cv = { 'method': 'GroupKFold', 'n_splits': 5, 'random_state': 42, 'shuffle': True, 'cv_target':'user' } IPHFQZ DWͷهࡌΛมߋ QZUIPOIPHFQZ GFBUVSFJNQPSUBODFʹ͍ͭͯ

Slide 60

Slide 60 text

࣮૷ৄࡉ ֶशͱಉ࣌ʹGFBUVSFJNQPSUBODF͕ը૾ϑΝΠϧͱͯ͠ग़ྗ͞ΕΔ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ

Slide 61

Slide 61 text

࣮૷ৄࡉ ֶशͱಉ࣌ʹGFBUVSFJNQPSUBODF͕ը૾ϑΝΠϧͱͯ͠ग़ྗ͞ΕΔ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ ͜ͷંΕઢάϥϑ͕ద੾Ͱ͸ͳ͍ͷ͸ঝ஌͓ͯ͠Γ·͢ʜ ࠓ͸ΦϨΦϨӡ༻ͳͷͰ͜ͷ··์ஔ͓ͯ͠Γ·ͯ͠ Ͳ͏͔ࢗͣ͞ʹ͓ئ͍͠·͢

Slide 62

Slide 62 text

࣮૷ৄࡉ ֶशͱಉ࣌ʹGFBUVSFJNQPSUBODF͕ը૾ϑΝΠϧͱͯ͠ग़ྗ͞ΕΔ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ ಛ௃ྔ

Slide 63

Slide 63 text

࣮૷ৄࡉ ֶशͱಉ࣌ʹGFBUVSFJNQPSUBODF͕ը૾ϑΝΠϧͱͯ͠ग़ྗ͞ΕΔ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ GPMEຖͷฏۉͱඪ४ภࠩ

Slide 64

Slide 64 text

࣮૷ৄࡉ ֶशͱಉ࣌ʹGFBUVSFJNQPSUBODF͕ը૾ϑΝΠϧͱͯ͠ग़ྗ͞ΕΔ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ มಈ܎਺ʢඪ४ภࠩฏۉʣ

Slide 65

Slide 65 text

࣮૷ৄࡉ ֶशͱಉ࣌ʹGFBUVSFJNQPSUBODF͕ը૾ϑΝΠϧͱͯ͠ग़ྗ͞ΕΔ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ ͜ͷ͋ͨΓͷಛ௃ྔ͸࡟ͬͯྑͦ͞͏ ͱ͍͏צॴ͕௫ΊΔ

Slide 66

Slide 66 text

࣮૷ৄࡉ ֶशͱಉ࣌ʹGFBUVSFJNQPSUBODF͕ը૾ϑΝΠϧͱͯ͠ग़ྗ͞ΕΔ ʙߴ଎ͳ1%$"Λ໨ࢦͯ͠ʙ ͜ͷ͋ͨΓͷಛ௃ྔ͸࡟ͬͯྑͦ͞͏ ͱ͍͏צॴ͕௫ΊΔ 1%$"଎͘ճͤͦ͏ײ

Slide 67

Slide 67 text

ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έͷ޻෉ʙ

Slide 68

Slide 68 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ

Slide 69

Slide 69 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ QMHC@@ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ

Slide 70

Slide 70 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ QMHC@@ 1VCMJDͷείΞ ʢʣ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ

Slide 71

Slide 71 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ QMHC@@ ֶशΛҰҙʹಛఆ͢Δ໊લ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ

Slide 72

Slide 72 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ QMHC@@ ΞϧΰϦζϜͷࣝผࢠ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ

Slide 73

Slide 73 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ QMHC@@ IPHFQZΛ࣮ߦͨ͠೔࣌ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ

Slide 74

Slide 74 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ

Slide 75

Slide 75 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ w ಛ௃ྔ w ύϥϝʔλ w GFBUVSFJNQPSUBODF w ֶशϩά w Ϟσϧ w ਪ࿦ϑΝΠϧ ͕อଘ͞ΕΔ

Slide 76

Slide 76 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ w ಛ௃ྔ w ύϥϝʔλ w GFBUVSFJNQPSUBODF w ֶशϩά w Ϟσϧ w ਪ࿦ϑΝΠϧ ͕อଘ͞ΕΔ ҙࣝͯ͠ͳ͍͚Ͳউखʹ؅ཧͰ͖ͯΔ

Slide 77

Slide 77 text

࣮૷ৄࡉ IPHFQZΛ࣮ߦͨ͋͠ͱͷϑΥϧμɾϑΝΠϧ ʙҙࣝ͠ͳͯ͘΋؅ཧͰ͖Δ࢓૊Έʙ w ಛ௃ྔ w ύϥϝʔλ w GFBUVSFJNQPSUBODF w ֶशϩά w Ϟσϧ w ਪ࿦ϑΝΠϧ ͕อଘ͞ΕΔ

Slide 78

Slide 78 text

·ͱΊ

Slide 79

Slide 79 text

·ͱΊ w ֶशɾਪ࿦ύΠϓϥΠϯ͍͍ͧʂ ύΠϓϥΠϯΛߏங͢Δ͜ͱͰԼهͷΑ͏ͳϝϦοτ͕͋Γʢࠓͷॴʣݸਓత ʹ͸ΊͬͪΌྑ͍ɻ ‣ ࠶ݱੑ ‣ ߴ଎ͳ1%$" ‣ ॾʑͷ؅ཧ w ্هͷΑ͏ͳ͜ͱ͕୲อ͞ΕΔͷͰ৺ཧత҆શੑ΋