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
データサイエンスで宝くじ hack
Search
tototti
August 02, 2016
Research
1.4k
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
データサイエンスで宝くじ hack
2016/8/2に、Closedな勉強会で発表した時の資料です。
「最近学んだデータサイエンスを利用して、宝くじで一儲けできないかな?」と考えてみた結果です。
tototti
August 02, 2016
More Decks by tototti
See All by tototti
ShazamKit調べてみた - 仙台iOS開発者勉強会 SWWDC47 - AppleEvent振り返り + LT会
tototti
0
250
2017年 SWWDCの活動振り返りと来年の展望
tototti
0
740
CoreMLやってみた - 2017/07/15 AKIBA.Swift x SWWDC
tototti
0
1.2k
Realm使ってみた (SWWDC 仙台iPhoneとか開発者勉強会 その26)
tototti
0
480
Xcodeの小ネタ (SWWDC 仙台iPhoneとか開発者勉強会 その26)
tototti
0
440
Other Decks in Research
See All in Research
OWASP AISVS - C7
shiell
2
760
[BlackHatAsia2026] Hidden Telemetry: Uncovering TraceLogging ETW Providers You're Not Using (Yet)
asuna_jp
1
710
CVPR2026論文紹介_VLMにとって良いvision encoderとは何か?Rethinking Model Selection in VLM Through the Lens of Gromov-Wasserstein Distance
kobayashi31
1
210
Karkada さんの論文 × 2 の紹介: (1) Closed-Form Training Dynamics Reveal Learned Features and Linear Structure in Word2Vec-like Models, (2) Symmetry in language statistics shapes the geometry of model representations
eumesy
PRO
1
740
JPA2026_NetworkTutorial_JunKashihara
junkashihara
0
150
クラウド・AI 時代の研究開発 DX / R&D Digital Transformation
hariby
0
140
医療LLMの現在地〜最新研究から社会実装までを考える〜
kento1109
1
1.7k
Visual SLAM未来予測 / Future Prediction in Visual SLAM
koide3
1
1k
HAKARI-Bench - 実運用視点での情報検索モデル評価ベンチマーク
hotchpotch
1
740
Physical AIでモデリングはどう変わるか/How Physical AI Will Change Modeling
stktu
0
120
[ACL 2026 Demo] Fast-MIA: Efficient and Scalable Membership Inference for LLMs
upura
0
120
Spatial Active Noise Control Based onSound Field Interpolation Incorporating Physical Constraints
skoyamalab
0
190
Featured
See All Featured
Neural Spatial Audio Processing for Sound Field Analysis and Control
skoyamalab
0
500
Raft: Consensus for Rubyists
vanstee
141
7.7k
Visualizing Your Data: Incorporating Mongo into Loggly Infrastructure
mongodb
49
10k
Lessons Learnt from Crawling 1000+ Websites
charlesmeaden
PRO
1
1.6k
AI Search: Where Are We & What Can We Do About It?
aleyda
0
7.9k
The Pragmatic Product Professional
lauravandoore
37
7.4k
How to Get Subject Matter Experts Bought In and Actively Contributing to SEO & PR Initiatives.
livdayseo
0
190
Why Mistakes Are the Best Teachers: Turning Failure into a Pathway for Growth
auna
0
280
The Cult of Friendly URLs
andyhume
79
7k
Tell your own story through comics
letsgokoyo
1
1.1k
Jess Joyce - The Pitfalls of Following Frameworks
techseoconnect
PRO
1
410
Keith and Marios Guide to Fast Websites
keithpitt
413
23k
Transcript
ๅ͘͡IBDL !UPUPUUJ σʔλαΠΤϯεͰ
ࣗݾհ w !UPUPUUJ w J04ΞϓϦ࡞ͬͯ·͢
ۙگ
૯ল౷ܭہͷແྉߨ࠲
σʔλαΠΤϯςΟετʹͳΓ·ͨ͠ ʁ
ֶΜͩ͜ͱΛཱ͍ͯͨ
ๅ͘͡IBDL
ๅ͘͡ w ී௨ͷๅ͘͡ w ൪߸ҹ͞Ε͍ͯΔɻ w ˠIBDLෆՄೳ w -050͘͡ w
൪߸ΛબΜͰങ͑Δɻ w ˠ)BDLՄೳ ͔
.*/*-050 w ʙ·Ͱͷࣈͷத͔ΒɺݸͷࣈΛબͿ w நͤΜࣈͱҰக͍ͯ͠Εɺબ
.*/*-050ͷظ ͨΓ݅ ֬ બֹۚ(ฏۉ) ظ 1 ਃࠐࣈ͕ຊࣈ5ݸͱશͯҰ கɻ 1
/ 169,911 10,046,800 59.13 2 ਃࠐࣈ5ݸͷ͏ͪ4ݸ͕ຊ ࣈʹҰக͠ɺͭΓ1ݸͷਃ ࠐࣈ͕ϘʔφεࣈͱҰ கɻ 5 / 169,911 144,300 4.25 3 ਃࠐࣈ5ݸͷ͏ͪ4ݸ͕ຊ ࣈʹҰகɻୠ͠2ͤΜ݅ Λআ͘ɻ 125 / 169,911 10,000 7.36 4 ਃࠐࣈ5ݸͷ͏ͪ3ݸ͕ຊ ࣈʹҰகɻ 3,250 / 169,911 1,000 19.13 ߹ܭ 89.86 ຕԁͳͷʹɺԁ͔͠ฦͬͯ͜ͳ͍൵͠Έʜ
rͤΜۚূථ๏ୈ lͤΜۚূථͷͤΜۚͷֹۚຢՁ֨ ͷ૯ֹɺͦͷൃച૯ֹͷޒׂʹ૬͢Δֹ Λ͑ͯ͜ͳΒͳ͍ɻz ͭ·Γɺ௨ৗๅ͘͡ͷظҎԼɻ
-050ͷબֹۚͷܾΊํ w બۚ༻ͷ༧ࢉΛɺબऀͰׂΔ w બऀ͕গͳ͍΄Ͳɺબֹ͕ۚଟ͍ w .*/*-050ͷ߹ɺ࠷େສԁ
.*/*-050 ઍສԁ ͨΓ݅ ֬ બֹۚ(ฏۉ) ظ 1 ਃࠐࣈ͕ຊࣈ5ݸͱશͯҰ கɻ
1 / 169,911 40,000,000 235.42 2 ਃࠐࣈ5ݸͷ͏ͪ4ݸ͕ຊ ࣈʹҰக͠ɺͭΓ1ݸͷਃ ࠐࣈ͕ϘʔφεࣈͱҰ கɻ 5 / 169,911 144,300 4.25 3 ਃࠐࣈ5ݸͷ͏ͪ4ݸ͕ຊ ࣈʹҰகɻୠ͠2ͤΜ݅ Λআ͘ɻ 125 / 169,911 10,000 7.36 4 ਃࠐࣈ5ݸͷ͏ͪ3ݸ͕ຊ ࣈʹҰகɻ 3,250 / 169,911 1,000 19.13 ߹ܭ 266.15 ຕԁͰɺԁͷϦλʔϯʜʂ
.*/*-050 ઍສԁ ͨΓ݅ ֬ બֹۚ(ฏۉ) ظ 1 ਃࠐࣈ͕ຊࣈ5ݸͱશͯҰ கɻ
1 / 169,911 30,000,000 176.56 2 ਃࠐࣈ5ݸͷ͏ͪ4ݸ͕ຊ ࣈʹҰக͠ɺͭΓ1ݸͷਃ ࠐࣈ͕ϘʔφεࣈͱҰ கɻ 5 / 169,911 144,300 4.25 3 ਃࠐࣈ5ݸͷ͏ͪ4ݸ͕ຊ ࣈʹҰகɻୠ͠2ͤΜ݅ Λআ͘ɻ 125 / 169,911 10,000 7.36 4 ਃࠐࣈ5ݸͷ͏ͪ3ݸ͕ຊ ࣈʹҰகɻ 3,250 / 169,911 1,000 19.13 ߹ܭ 207.29 ສԁͰɺظϓϥεɻ
.*/*-050 ೋઍສԁ ͨΓ݅ ֬ બֹۚ(ฏۉ) ظ 1 ਃࠐࣈ͕ຊࣈ5ݸͱશͯҰ கɻ
1 / 169,911 20,000,000 117.71 2 ਃࠐࣈ5ݸͷ͏ͪ4ݸ͕ຊ ࣈʹҰக͠ɺͭΓ1ݸͷਃ ࠐࣈ͕ϘʔφεࣈͱҰ கɻ 5 / 169,911 144,300 4.25 3 ਃࠐࣈ5ݸͷ͏ͪ4ݸ͕ຊ ࣈʹҰகɻୠ͠2ͤΜ݅ Λআ͘ɻ 125 / 169,911 10,000 7.36 4 ਃࠐࣈ5ݸͷ͏ͪ3ݸ͕ຊ ࣈʹҰகɻ 3,250 / 169,911 1,000 19.13 ߹ܭ 148.44 ສԁͩͱɺظϚΠφεɻ
ͭ·Γʜ w ߪೖऀ͕ଟ͍ɺ୭͕ങ͏Α͏ͳ બֹ͕ۚ গͳ͍ ߹ͤΛߪೖ͢Δͱɺ࣮֬ʹଛΛ͢Δ w ߪೖऀ͕গͳ͍ બֹ͕ۚଟ͍ ߹ͤΛങ͏
͜ͱͰɺऩࢧΛϓϥεʹग़དྷΔՄೳੑ͕͋Δɻ
աڈͷબֹۚ ສҎ্ଘࡏ͍ͯ͠Δɻ
աڈͷબֹۚ ສҎ্͕ଘࡏʜ ֹۚ ʙ ສʙ ສʙ ສʙ
ສʙ ສʙ ສʙ ສʙ ສ ׂ߹
΅͘ͷ͔Μ͕͑ͨ ͍͖͞ΐ͏ͷઓུ w .*/*-050ͷ߹ͤ ͷ͏ͪɺظ ͕ϓϥεͷͷΛʰશͯʱങ͏ w ͕ສԁҎ্ͱͳΔ߹ͤͷΈߪೖ
w શମͷ ௨Γສԁ ͕ࣗങ͏બ͕ۚԼ͕ΔͷͰɺϚʔδϯऔͬͯ·͢
΅͘ͷ͔Μ͕͑ͨ ͍͖͞ΐ͏ͷઓུ w ظ͕ϓϥεͷ͘͡Λങ͍ଓ͚Δ w ظ͕ϓϥεͰ͋Εɺࢼߦճ͕ଟ͍΄ Ͳϓϥεʹऩଋ͍ͯ͘͠ w ສԁߪೖΛؒ ճ
ଓ͚Εɺ ˋͰ͕ͨΔˠऩࢧ͕ʴʹͳΔ ͷ͘͡ΛճҾ͍ͨ࣌ɺճҎ্ͨΔ֬㲈 F
બֹۚ༧ଌ
εΫϨΠϐϯά w աڈͷ.*/*-050ͷ݁ՌΛεΫϨΠϐϯάͰऔಘɻ w $ISPNFͷϓϥάΠϯͱ͔Λ͍·ͨ͠ɻ ճͷσʔλΛऩू͠ɺ$47ʹ͠·ͨ͠ɻ
աڈͷສԁύλʔϯ w w
w w w w ͳΜ͔ɺ ͋ͨΓ͕ո͍͠ʜ
ॏճؼੳ w ॏճؼੳʢ͡Ύ͏͔͍͖ͿΜ͖ͤʣɺଟม ྔղੳͷҰͭɻճؼੳʹ͓͍ͯಠཱม͕ͭ Ҏ্ʢ࣍ݩҎ্ʣͷͷɻ<8JLJQFEJB> ͬͱྑ͍ख๏͕͋Γͦ͏ͳؾ͕͠·͢ʜ
ࣜ • y = a1ɾx1 + a2ɾx2 + … anɾxn
+ b w ैଐมZ w બऀͷׂ߹ w આ໌มY w ʙͷͲͷࣈ͕બ͞Ε͔ͨʁ w ࿈ଓࣈ͕બ͞Ε͔ͨʁ
w 3ݴޠʢ͋ʔΔ͛Μ͝ʣΦʔϓϯιʔεɾϑϦʔ ιϑτΣΞͷ౷ܭղੳ͚ͷϓϩάϥϛϯά ݴޠٴͼͦͷ։ൃ࣮ߦڥͰ͋Δɻ 8JLJQFEJB &YDFMͰग़དྷΔΜͰ͕͢ɺIBDLײΛग़͍ͨ͠ͷͰʜ
demo
݁Ռੳ֤ࣈͷਓؾ ෆਓؾͳͷ ͋ͨΓ
݁Ռੳͦͷଞ w ࿈ଓͨ͠ࣈ ͱ͔ ͕ೖ͍ͬͯΔͱɺ બ্͕͕ۚΔ w ࿈ଓݸ·Ͱͦ͏͕ͩɺ࿈ଓݸҎ্ʁ
w Ұͭඈ͠ ͱ͔ ΄΅ಉ͡ ࿈ଓݸҎ্σʔλ͕গͳ͍Өڹ͋Γʜɻ
݁Ռݕূ w NJOJMPUPୈճ w ɹɹɹɹ ԁ w ݁Ռ༧ଌʜ w
ສԁ w ສԁ ສʙສͷൣғͱ༧ଌग़དྷͯΔʜ
݁Ռੳ w ༧͕ͨΔ߹͋Δ w ͔͠͠೦ͳ͕Βɺ༧ਫ਼͍ w BEKVTUFE3ͷ͕ˠྑ͍༧ଌͱ͍͑ ͳ͍ w આ໌มͷՃ͕ඞཁͱࢥΘΕΔ
·ͱΊ w ๅ͘͡ ˞ݱ࣮ੑͱ͔͘ IBDLՄೳ w ˞ͨͩ͠ɺϩτ͘͡ʹݶΔ w ˞ͨͩ͠ɺબۚΛ༧ଌͰ͖Δ߹ʹݶΔ w
˞ͨͩ͠ɺଞͷߪೖऀ͕ಉ͡ख๏ΛΘͳ͍ ߹ w ˞ͨͩ͠ɺڊֹͷࢿۚΛࣦͬͯྑ͍߹
ࠓޙͷ՝ w બۚ༧ଌਫ਼ͷ্ w .BDIJOF-FBSOJOHͱ͔ͰͰ͖ͳ͍ͷ͔ʜ w ͬͱ౷ܭͷֶश͕ඞཁʜ w બֹۚ༧ΞϓϦͱ͔࡞ͬͯΈ͍ͨ