Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Sign up for free
Menu
Search
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Pricing
Search
Sign in
Sign up for free
データサイエンスLT祭 のLTネタ
Search
tetsuroito
September 07, 2016
1.9k
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
データサイエンスLT祭 のLTネタ
Low Bias High VarianceとFilter Bubbleな私
tetsuroito
September 07, 2016
More Decks by tetsuroito
See All by tetsuroito
白金鉱業meetup発表資料
tetsuroito
1
480
Data Engineering Study#30 LT資料
tetsuroito
2
2.3k
データエンジニアリングの潮流を俯瞰する
tetsuroito
1
2k
Classiが取り組んできた 機械学習の試行錯誤
tetsuroito
0
930
事業会社でのデータマネジメントのプラクティス #TechMar
tetsuroito
1
730
Data Engineering Study #9 Classiのデータ組織の歩み
tetsuroito
5
6.4k
Data Engineering Study #3 基調講演_データ分析基盤の浸透に必要なこと
tetsuroito
4
5.4k
Subscription Meetup Vol.2 Opening Talk Slide
tetsuroito
0
200
Data_Pipeline_Casual_Talk_Vol.4_for_Ready.pdf
tetsuroito
0
1.8k
Featured
See All Featured
Why Your Marketing Sucks and What You Can Do About It - Sophie Logan
marketingsoph
0
470
Neural Spatial Audio Processing for Sound Field Analysis and Control
skoyamalab
0
610
What's in a price? How to price your products and services
michaelherold
247
13k
Save Time (by Creating Custom Rails Generators)
garrettdimon
PRO
33
5.7k
The untapped power of vector embeddings
frankvandijk
2
2k
A Guide to Academic Writing Using Generative AI - A Workshop
ks91
PRO
1
560
Building Flexible Design Systems
yeseniaperezcruz
330
41k
How to train your dragon (web standard)
notwaldorf
96
6.9k
How to make the Groovebox
asonas
2
2.5k
Docker and Python
trallard
47
4.3k
The #1 spot is gone: here's how to win anyway
tamaranovitovic
4
1.2k
Why Mistakes Are the Best Teachers: Turning Failure into a Pathway for Growth
auna
0
350
Transcript
Low Bias High Variance ͱFilter Bubbleͳࢲ Datascience LTࡇ @tetsuroito 2016/09/07ɹɹYahoo
Japan @Tokyo Middtown
ࣗݾհ ໊લɿҏ౻ ప (@tetsuroito) ࣄɿاըͱ͔ੳͱ͔ࡶ༻पΓ ϚΠϒʔϜɿᰤࢠ ҙؾࠐΈɿ͓ۚΛલʂ ͔͔ͤͬͩ͘Β͍͔ͭ͘એ͍ͨ͠
ZDNetͰϙΤϜॻ͍ͯ·͢
ZDNetͰϙΤϜॻ͍ͯ·͢ ΈΜͳͷγΣΞͯϒ͕ ࢲͷϞνϕʔγϣϯͰ͢ʂ
Tγϟπ࡞ͬͨ
Tγϟπ࡞ͬͨ པΉങ͍ͬͯͩ͘͞w https://steers.jp/c/whyareuusingsjis
ؓٳ
ຊͷΞδΣϯμ Low Bias High Varianceͳ Filter Bubbleͳ ͦͯ͠ࢲ
Low Bias High Varianceͳ σʔλղੳͷత →ݱͷഎܠʹ͋Δ͘͠Έͷಛఆ ɹۙࣅతʹସՄೳͳ౷ܭϞσϧͷߏங σʔλղੳͷ࣮ࡍ →ͨ·ͨ·ಘΒΕͨσʔλͷͯ·Γ ɹ౷ܭϞσϧͷෳࡶԽ
Ͳ͏ͯ͜͠͏ͳͬͯ͠·͏ͷ͔ ؍ଌσʔλ͕KPIͷηϯγςΟϏςΟ͕ߴ͍ KPIΛΉํͲΜͲΜաֶश͍ͯ͘͠ ࣮ͷݱͰ ͜ͷΑ͏ͳ౷ܭϞσϦϯά͕ࢍ͞ΕΔ
ྫ͑ɺ͜Μͳࢪࡦͷ۩߹ tظͷࢪࡦ ධՁ× ධՁ˓ ධՁ˚ t+1ظͷࢪࡦ ଧͪΓ ܧଓ ܧଓ New
ଧͪΓ ੜଘόΠΞε͕࣌ܥྻ૬্͕͍ؔͯͬͯ͘͠ t+2ظͷࢪࡦ
͏·͍ͬͨ͘ࢪࡦ… ޮ༻ۂઢͷݟ͔ΒɺঃʑʹޮՌ͕ബΕΔ ޮ༻ਫ४ ࣮ࢪճ ޮ༻ۂઢ ཧ্ɺࠨهͷΑ͏͕ͩɺ ͔͢͠Δͱ͖ϚΠφε͔
ฏۉճؼͷ᠘ ॳ͍όΠΞε ߴ͍όϦΞϯεͰ ࢪࡦΛ࣮ࢪ ϓϩηεΛܦΔʹͭΕͯɺ ͲΜͲΜฏۉʹऩଋ
γϛϡϨʔγϣϯ͕େࣄͰ͢Ͷ ϒʔτετϥοϓ๏ ϞϯςΧϧϩγϛϡϨʔγϣϯ ͳͲͷΠςϨʔγϣϯճΛ ճͯ͠ऩଋ͢ΔϞσϦϯά ͕ΩʔͱͳΔ͔
Filter Bubbleͳ ݟ͍ͨͷ͔͠Έͳ͘ͳΔΑ͏ʹͳΔ
όϦΞϯε όΠΞε ϑΟϧλʔόϒϧ
ใߑਫͷӔ ใͷऔࣺબඞࢸ
ωοτόΧʹֶͿ χίϥεɾΧʔᐌ͘ Googleͷݕࡧߦಈ͕ͷใुܥʹϑΟʔυόο ΫΛ༩͍͑ͯΔɻ ݕࡧˠ͑ͷڧԽֶश ҙྗ͕ࢄອʹͳΓɺਂ͍ࢥߟ͕Ͱ͖ͳ͘ͳΔ
εΫϥοϓ&Ϗϧυ͕ඞཁͩʂ ͜Ε·ͰͷܦݧΛյͯ͠ ৽͍͠ϞσϧΛߏங
తഁյͬͯ ͦ͏͍͏͜ͱʁ γϡϯϖʔλʔ͞Μ
でも、現実には難しい だって人間だもの てつろー
ͦΜͳ͜ͱΛߟ͑ͳ͕Β ʑͷੳاըͳͲΛ ߟ͍͑ͯ·͢ ົҊ͋Εڭ͑ͯԼ͍͞ʂ
ご静聴 ありがとうございました