Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Data Science BOOTCAMP Practices - Recommendation
Search
Yohei Munesada
May 09, 2017
Technology
0
170
Data Science BOOTCAMP Practices - Recommendation
レコメンデーションの制作演習のスライドです。中に解答例のリンクも掲載しています。
G's Academy Data Science Bootcamp
Yohei Munesada
May 09, 2017
Tweet
Share
More Decks by Yohei Munesada
See All by Yohei Munesada
G'sデータベース設計の講義
yoheimune
4
5k
How to create a service, How to google !
yoheimune
0
270
Machine Learning Basic and Python
yoheimune
1
440
Python Scraping and Web Apps for G's ACADEMY TOKYO
yoheimune
0
210
DevelopWorkflow and Solving Problems
yoheimune
0
400
Git and Github for Beginners
yoheimune
1
250
Data Science BOOTCAMP Practices
yoheimune
0
340
Machine Learning with Python
yoheimune
0
300
Python Basics for G's ACADEMY TOKYO
yoheimune
1
520
Other Decks in Technology
See All in Technology
Ask-LLM論文紹介: How to Train Data-Efficient LLMs
s_ota
0
110
PG-Stromの性能評価レポート~ Star Schema Benchmark を例に~ / pgstrom_ssb_report_2024
sakaik
0
100
沒想過的前端錯誤處理可能比你有做的還多
line_developers_tw
PRO
0
1.8k
匠MethodとRDRAとICONIXとDDDで実現する一気通貫オブジェクト指向開発
haru860
4
1.5k
オーティファイ会社紹介資料 / Autify Company Deck
autifyhq
7
100k
TCA入門したてなので、自分が馴染みのある実装と比較しながらキャッチアップしてみる
fumiyasac0921
1
370
なんで私に登壇依頼が?! ~頼られるエンジニアになるためには~ /
mixi_engineers
PRO
2
200
家族アルバム みてねで直面してきた技術的負債 / MIXI KAG 2024
isaoshimizu
17
7.7k
Azureコストは水道代/The_47th_Tokyo_Jazug
aeonpeople
3
350
今さら聞けない!? AWSの生成AIサービス Amazon Bedrock入門!
minorun365
PRO
11
1.8k
サービス成長と共に肥大化するモノレポ、長くなるCI時間 / As services grow, monorepos get bigger and CI time gets longer
kohbis
5
2k
AMLD 2024 - Build Your Own GPT
donlelef
1
260
Featured
See All Featured
Product Roadmaps are Hard
iamctodd
43
9.6k
How to Ace a Technical Interview
jacobian
272
22k
Atom: Resistance is Futile
akmur
258
25k
Rebuilding a faster, lazier Slack
samanthasiow
72
8.2k
Producing Creativity
orderedlist
PRO
335
39k
A Philosophy of Restraint
colly
195
15k
The Success of Rails: Ensuring Growth for the Next 100 Years
eileencodes
28
5.9k
BBQ
matthewcrist
78
8.7k
ParisWeb 2013: Learning to Love: Crash Course in Emotional UX Design
dotmariusz
101
6.6k
RailsConf 2023
tenderlove
0
510
Why Our Code Smells
bkeepers
PRO
330
56k
Templates, Plugins, & Blocks: Oh My! Creating the theme that thinks of everything
marktimemedia
18
1.7k
Transcript
Data Science BOOTCAMP Ϩίϝϯσʔγϣϯ࡞ Yohei Munesada
About Me 㾎फఆ༸ฏ ΉͶͩ͞Α͏͍ 㾎 ג αΠόʔΤʔδΣϯτ 㾎(`TΞΧσϛʔϝϯλʔ 㾎IUUQXXXZPIFJNOFU 㾎ͱσʔλαΠΤϯε
Time tables 19:30ʙ19:40ɹΦʔϓχϯάͱࠓͷׂ࣌ؒ 19:40ʙ19:50ɹάϧʔϓϫʔΫઆ໌ 19:50ʙ20:30ɹάϧʔϓϫʔΫʢൃද४උʣ 20:30ʙ20:40ɹٳܜ 20:40ʙ21:30ɹάϧʔϓผൃදʢ5 x 7νʔϜ +
αʣ 21:30ʙ21:40ɹ࣍ͷ՝ͷઆ໌ʢ͞Βͬͱʣ 21:40ʙ21:50ɹάϧʔϓϫʔΫʢऔΓΈ༰ͷڞ༗ͱϒϥογϡΞοϓʣ 21:50ʙ22:00ɹऔΓΈ༰ͷൃදʢ30ඵ x 7νʔϜ + αʣ
Exercises - MovieLens .PWJF-FOTΛ༻͍ͨϨίϝϯσʔγϣϯͷߏங ඞਢ՝ .PWJF-FOTͱ͍͏ެ։σʔλʹɺөըͷใɺϢʔβʔͷөըʹର͢Δใ ͳͲؚ͕·Ε·͢ɻͦΕΒσʔλΛ༻͍ͯϨίϝϯυγεςϜΛߏங͍ͯͩ͘͠͞ɻ ٻΊΔΞτϓοτ ɹɾϢʔβʔʹରͯ͠өըΛਪન͢Δ
ϙΠϯτ ɹɾਪનʹ͍ͭͯͲͷΑ͏ʹػցֶशͱͯ͠ఆٛ͢Δ͔ʁ ɹɾͳͥͦͷϞσϧΛબ͢Δͷ͔ʁ ɹɾ༧ଌ݁ՌͷධՁ݁ՌʁͲͷΑ͏ʹධՁ͢Εྑ͍͔ʁ
Exercises - MovieLens
Presentation contents ʢՄೳͰͨ͠ΒʣσϞ ͲͷΑ͏ͳػցֶशͱͯ͠ఆ͔ٛͨ͠ʁ ͲͷΑ͏ͳ࣮Λ͔ͨ͠ʁ ͲͷΑ͏ʹϞσϧΛධՁ͔ͨ͠ʁ
ͨ͠ͱ͜Ζɺۤ࿑ͨ͠ͱ͜Ζ ͦͷଞओு͍ͨ͜͠ͱΛͲ͏ͧʂ
Group work ݸਓͰͷՌΛνʔϜͰൃද͢Δ νʔϜͱͯ͠ͷൃද༰Λ࡞͢ΔʢϓϨθϯܗࣜࣗ༝ʣ άϧʔϓϫʔΫΛߦ͍·͢ ʢʙʣ ʢՄೳͰͨ͠ΒʣσϞ
ͲͷΑ͏ͳػցֶशͱͯ͠ఆ͔ٛͨ͠ʁ ͲͷΑ͏ͳ࣮Λ͔ͨ͠ʁ ͲͷΑ͏ʹϞσϧΛධՁ͔ͨ͠ʁ ͨ͠ͱ͜Ζɺۤ࿑ͨ͠ͱ͜Ζ ͦͷଞओு͍ͨ͜͠ͱΛͲ͏ͧʂ ϓϨθϯ༰
Take a break ͓ർΕ༷Ͱͨ͠ɺٳܜͰ͢ ʢʙʣ
How is your recommend system ? ൃදͷ͓࣌ؒͰ͢ʂ
How is your recommend system ? ղྫ https://goo.gl/4jGdHI
Next exercises .PWJF-FOTΛ༻͍ͨϨίϝϯσʔγϣϯͷߏங ࠃௐࠪσʔλΛ༻͍ͨσʔλαΠΤϯε ҙͷެ։σʔλΛ༻͍ͨػցֶश ػցֶशܥΫϥυ"1*Λ༻͍ͨαʔϏε։ൃ
ඞਢ՝ બ՝
Next exercises - ࠃௐࠪ ࠃௐࠪσʔλΛ༻͍ͨσʔλαΠΤϯε બ՝ ࠃௐࠪσʔλ͔ΒਓޱɺՈߏɺ৬ۀͳͲ༷ʑͳใΛಘΔ͜ͱ͕Ͱ͖·͢ɻ ԿΒ͔ͷϏδωε՝Λఆٛͨ͠ͷͪʹɺࠃௐࠪσʔλΛ༻͍ͯϏδωεͷ ҙࢥܾఆΛॿ͚ΔใΛఏ͍ࣔͯͩ͘͠͞ɻ
ٻΊΔΞτϓοτ ɹɾఆٛͨ͠Ϗδωε՝Կ͔ʁ ɹɾͦΕʹରͯ͠ࠃௐࠪσʔλΛͲͷΑ͏ʹ׆༻͔ͨ͠ʁ Ϗδωε՝ྫ ɹɾ*5ڭҭϏδωεΛల։͍ͨ͠ɻͲͷࢢொଜΛλʔήοτʹ͢Δ͖͔ʁ ɹɾϑΟϦϐϯਓʹ͚ͨΧϑΣϏδωεΛߦ͍͍ͨɻͲ͜ͰΔ͔ʁ ɹɾͳͲ
ར༻Մೳͳσʔλ ɹIUUQXXXTUBUHPKQEBUBLPLVTFJJOEFYIUN Next exercises - ࠃௐࠪ ࠃௐࠪσʔλΛ༻͍ͨσʔλαΠΤϯε બ՝
Next exercises - ࠃௐࠪ
Next exercises - ҙͷσʔλͰʂ ҙͷެ։σʔλΛ༻͍ͨػցֶश બ՝ ੈͷதʹ༷ʑͳσʔλ͕ެ։͞Ε͓ͯΓɺػցֶशʹར༻Ͱ͖Δσʔλ ଟʑଘࡏ͠·͢ɻڵຯͷ͋Δσʔλʹ͍ͭͯԾઆΛఆٛͯ͠ػցֶशΛߦ͍ɺ ԿΒ͔ͷՌΛग़͢औΓΈΛ͍ͯͩ͘͠͞ɻ
ٻΊΔΞτϓοτ ɹɾͲͷΑ͏ͳσʔλΛ͏͔ʁ ɹɾͲΜͳԾઆΛઃఆ͔ͨ͠ʁ ɹɾͲͷΑ͏ͳՌΛಋ͍ͨͷ͔ʁ·ͨͦΕΛͲͷΑ͏ʹಋ͍ͨͷ͔ʁ
ར༻Մೳͳσʔλྫ ɹ6$*.BDIJOF-FBSOJOH ɹɹIUUQBSDIJWFJDTVDJFEVNM ɹࠃཱใֶݚڀॴ ɹɹIUUQXXXOJJBDKQETDJESEBUBMJTUIUNM ɹ%"5"(0+1 ɹɹIUUQXXXEBUBHPKQ ɹ*NBHF/FU ɹɹIUUQXXXJNBHFOFUPSH Next
exercises - ҙͷσʔλͰʂ ɹ,BHHMF ɹɹIUUQTXXXLBHHMFDPNEBUBTFUT ɹ-JWFEPPSχϡʔε ɹɹIUUQOFXTMJWFEPPSDPN ɹ౦ژϝτϩΦʔϓϯσʔλ ɹɹIUUQTEFWFMPQFSUPLZPNFUSPBQQKQJOGP ɹ5XJUUFS"1*ɺͳͲ ҙͷެ։σʔλΛ༻͍ͨػցֶश બ՝
Next exercises - ҙͷσʔλͰʂ
Next exercises - ػցֶशAPIΛͬͯʂ ػցֶशܥΫϥυ"1*Λ༻͍ͨαʔϏε։ൃ બ՝ (PPHMF"84"[VSF#JOH*#.ͷ֤αʔϏεͰػցֶशܥͷ"1*͕ ఏڙ͞Ε͍ͯΔʢྫɿإೝࣝɺԻೝࣝɺςΩετUPεϐʔνɺFUDʣɻ ͜ΕΒͷ"1*Λ͍ɺԿΒཱ͔ͪͦ͏ͳΞϓϦαʔϏεΛ੍࡞͍ͯͩ͘͠͞ɻ
ٻΊΔΞτϓοτ ɹɾͲͷ"1*Λར༻͢Δͷ͔ʁ ɹɾԿʹཱͯΔͷ͔ʁͲͷΑ͏ͳαʔϏε͔ʁ ग़ҙਤ ɹɾֶशࡁΈͷϞσϧΛͲͷΑ͏ʹ࣮ੈքͰ׆͔͢ͷ͔ɺͦΕΛߟ͑ߦಈ͢Δɻ
Next exercises - ػցֶशAPIΛͬͯʂ
Next exercises .PWJF-FOTΛ༻͍ͨϨίϝϯσʔγϣϯͷߏங ࠃௐࠪσʔλΛ༻͍ͨσʔλαΠΤϯε ҙͷެ։σʔλΛ༻͍ͨػցֶश ػցֶशܥΫϥυ"1*Λ༻͍ͨαʔϏε։ൃ
ඞਢ՝ બ՝
Group work ݸਓͦΕͧΕͰऔΓΜͰ͍Δ༰ʢऔΓΉ༰ʣΛڞ༗ ൃද༰·ͱΊʢϓϨθϯܗࣜޱ಄Ͱʣ άϧʔϓϫʔΫΛߦ͍·͢ ʢʙʣ
Group work ൃදʢͲͷΑ͏ͳ༰Λѻ͏͔ʣ άϧʔϓϫʔΫΛߦ͍·͢ ʢʙʣ
Thank you ͦΕͰྑ͍σʔλαΠΤϯεΛʂ