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
これからの強化学習2.7
Search
moyomot
May 19, 2017
0
110
これからの強化学習2.7
moyomot
May 19, 2017
Tweet
Share
More Decks by moyomot
See All by moyomot
DRIVE CHARTのMLOpsを体感しよう
moyomot
0
51
現場課題に向き合い MLOps成熟度を高める道
moyomot
1
830
第1回 Data-Centric AI勉強会 LT: AIドラレコを支える一貫性のあるデータの作り方
moyomot
0
680
DRIVE CHARTにおけるAI開発とアーキテクチャ全容
moyomot
0
650
これからの強化学習2.6
moyomot
0
190
Gunosyのデータ分析基盤、ログ基盤の全容
moyomot
14
9.2k
GunosyにおけるSparkStreaming活用事例
moyomot
0
4.9k
トピックモデル第2章
moyomot
0
280
adhoc analysis apache spark
moyomot
1
1k
Featured
See All Featured
Creating an realtime collaboration tool: Agile Flush - .NET Oxford
marcduiker
13
1.5k
Designing the Hi-DPI Web
ddemaree
276
33k
How to train your dragon (web standard)
notwaldorf
72
5.1k
Optimising Largest Contentful Paint
csswizardry
7
2.3k
Exploring the Power of Turbo Streams & Action Cable | RailsConf2023
kevinliebholz
1
3.4k
Done Done
chrislema
178
15k
What the flash - Photography Introduction
edds
64
11k
Keith and Marios Guide to Fast Websites
keithpitt
408
22k
How STYLIGHT went responsive
nonsquared
92
4.8k
Refactoring Trust on Your Teams (GOTO; Chicago 2020)
rmw
24
2.3k
WebSockets: Embracing the real-time Web
robhawkes
59
7k
The Illustrated Children's Guide to Kubernetes
chrisshort
29
46k
Transcript
͜Ε͔ΒͷڧԽֶश 2.7 ෳརܕڧԽֶश GUNOSY σʔλϚΠχϯάݚڀձ #121
INTRODUCTION ͓͢Δ͜ͱ ▸ རӹͷෳརޮՌΛ۩ମྫΛ௨ͯ͠ཧղ͠ ▸ ෳརΛQֶशͷΈʹఆࣜԽ͢Δ
INTRODUCTION ࣍ ▸ 2.7.1 རӹͷෳརޮՌͱࢿൺ ▸ زԿฏۉΛ༻ͨ͠ෳརޮՌͷ۩ମྫ ▸ 2.7.2 ෳརܕڧԽֶशͷΈ
▸ ߦಈՁؔQͷఆࣜԽ ▸ 2.7.3 ෳརܕڧԽֶशΞϧΰϦζϜ ▸ ෳརܕQֶश ▸ ෳརܕOnPSʢOnline Profit Sharingʣ ▸ 2.7.4 ࢿൺͷ࠷దԽ ▸ 2.7.5 ϑΝΠφϯεͷԠ༻ɿࠃ࠴ฑબ
2.7.1 རӹͷෳརޮՌͱࢿൺ 3ຊόϯσΟοτ ▸ ͲͷϚγϯ͕͓ಘ͔ʁ
2.7.1 རӹͷෳརޮՌͱࢿൺ ͲͷϚγϯ͕͓ಘ͔ʁ ▸ ͷຊ࣭ࢉज़ฏۉ͔زԿฏۉ͔ ▸ ֫ಘͨ͠རӹؚΊͯશֹ͔͚ଓ͚Δͱ͖زԿฏۉΛߟྀ͢Δ ඞཁ͕͋ΔʢAͷબ͕ྑ͍ʣ ▸ زԿฏۉෳརͷΑ͏ͳൺͰมԽ͢Δͱ͖ʹ༻͢Δ
▸ ʢຖճ1υϧBET͢Δ߹ࢉज़ฏۉͷ΄͏͕ྑ͍݁ՌʹͳΔ ͣʣ ▸ https://www.jstage.jst.go.jp/article/tjsai/26/2/26_2_330/_pdf
2.7.1 རӹͷෳརޮՌͱࢿൺ ෳརޮՌΛ࠷େԽ͢ΔͨΊʹέϦʔج४ ▸ ΫϩʔυɾγϟϊϯΒͱڞʹɺใڞ༗Λར༻͠ɺΪϟϯϒϧͰ࠷ޮͷΑ ͍Ṍ͚ํΛݚڀͨ͠ɻͨͩ͠ɺࣗΒṌ͚Δ͜ͱ͠ͳ͔ͬͨɻʢwikipediaʣ
2.7.2 ෳརܕڧԽֶशͷΈ ऩӹׂҾͷൺֱ ऩӹͷׂҾ ׂҾෳརརӹ R: རӹ, γ: ׂҾ, f:
ࢿൺ r: ใु త: ߦಈՁؔQΛ࠶ؼతͳܗͰఆࣜԽ͠Q- learningʹ͍͖͍࣋ͬͯͨ
2.7.2 ෳརܕڧԽֶशͷΈ ঢ়ଶՁ؍ͷఆࣜԽ
2.7.2 ෳརܕڧԽֶशͷΈ ߦಈՁ؍ͷఆࣜԽ ͋ͱQΛ࠷େԽ͢ΔํࡦπΛֶश͢Δ
2.7.3 ෳརܕڧԽֶशΞϧΰϦζϜ ෳརܕQֶश ▸ ҰൠతͳQֶशͱߟ͑ํಉ͡ ▸ ใुΛརӹͷରͰஔ͖͑ͨ
2.7.3 ෳརܕڧԽֶशΞϧΰϦζϜ ෳརܕQֶशͷΞϧΰϦζϜ
2.7.3 ෳརܕڧԽֶशΞϧΰϦζϜ ෳརܕOnPS(ONLINE PROFIT SHARING) ▸ Profit Sharing ▸ QֶशϚϧίϑੑɺProfit
SharingඇϚϧίϑੑOK ▸ ঢ়ଶs, ߦಈaͷ༏ઌΛPΛஔ͘, F৴༻ׂؔʢڧԽؔ ʣ ▸ ใु֫ಘʹෆඞཁͳߦಈΛଟ࣮͘ߦ͢Δඇ߹ཧͳํࡦΛֶ श͢Δ՝͋Γ ▸ https://www.jstage.jst.go.jp/article/fss/27/0/27_0_304/ _pdf
2.7.3 ෳརܕڧԽֶशΞϧΰϦζϜ ෳརܕOnPSͷΞϧΰϦζϜ
2.7.4 ࢿൺͷ࠷దԽ ࢿൺGͬͯͲ͏ͬͯબͿͷʁ ▸ ΦϯϥΠϯޯ๏ͰfΛߋ৽͢ΕOK
Ͳͷࠃͷࠃ࠴Λߪೖ͢ΕΑ͍͔ ▸ ෳརܕQֶशͱैདྷͷQֶशΛൺֱ ▸ ෳརܕQֶशزԿฏۉͰརӹͷ ෳརޮՌΛେ͖͘͢Δ͜ͱ͕Ͱ͖ͯ ͍Δ 2.7.5 ϑΝΠφϯεͷԠ༻ྫɿࠃ࠴ฑબ