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
Deep Learning Book 10その2 / deep learning book 1...
Search
Sponsored
·
Your Podcast. Everywhere. Effortlessly.
Share. Educate. Inspire. Entertain. You do you. We'll handle the rest.
→
himkt
January 29, 2018
Research
210
2
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
Deep Learning Book 10その2 / deep learning book 10 vol2
himkt
January 29, 2018
More Decks by himkt
See All by himkt
Linformer: paper reading
himkt
0
640
RoBERTa: paper reading
himkt
1
420
NLP SoTA 勉強会 / ner_2019
himkt
2
1.5k
自然言語処理 @ クックパッド / nlp at cookpad
himkt
1
580
Interpretable Machine Learning 6.3 - Prototypes and Criticisms
himkt
2
200
ニューラル固有表現抽出 / Neural Named Entity Recognition
himkt
3
840
ニューラル固有表現抽出器を実装してみる / PyNER
himkt
6
2.2k
Spacyでお手軽NLP / NLP with spacy
himkt
0
1.1k
ふわふわ系列ラベリング / ner 2018
himkt
5
880
Other Decks in Research
See All in Research
第12回人と環境にやさしい交通をめざす全国大会/熊本都市圏「車1割削減、渋滞半減、公共交通2倍」をめざして
trafficbrain
0
110
AIを叩き台として、 「検証」から「共創」へと進化するリサーチ
mela_dayo
0
280
NLP colloquium: AI Safety Survey
kanekomasahiro
0
540
姫路市 -都市OSの「再実装」-
hopin
0
1.7k
適応的スパムフィルタのための軽量な類似メッセージカウンタ / jsai2026-adaptive-spam-filter
monochromegane
0
2.1k
AIエージェント時代のLLM-jpモデルのあるべき姿
k141303
0
440
非試合日の野球場を楽しむためのARホームランボールキャッチ体験システムの開発 / EC79-miyazaki
yumulab
0
210
英語教育 “研究” のあり方:学術知とアウトリーチの緊張関係
terasawat
1
990
正規分布と最適化について
koide3
1
240
LOSの検討(λ Kansai 2026 in Winter)
motopu
0
140
National high-resolution cropland classification of Japan with agricultural census information and multi-temporal multi-modality datasets
satai
3
270
2026 東京科学大 情報通信系 研究室紹介 (大岡山)
icttitech
0
3.7k
Featured
See All Featured
Java REST API Framework Comparison - PWX 2021
mraible
34
9.3k
Building an army of robots
kneath
306
46k
The World Runs on Bad Software
bkeepers
PRO
72
12k
Data-driven link building: lessons from a $708K investment (BrightonSEO talk)
szymonslowik
1
1.1k
Fight the Zombie Pattern Library - RWD Summit 2016
marcelosomers
234
17k
The Art of Programming - Codeland 2020
erikaheidi
57
14k
VelocityConf: Rendering Performance Case Studies
addyosmani
333
25k
Responsive Adventures: Dirty Tricks From The Dark Corners of Front-End
smashingmag
254
22k
A better future with KSS
kneath
240
18k
XXLCSS - How to scale CSS and keep your sanity
sugarenia
250
1.3M
Reality Check: Gamification 10 Years Later
codingconduct
0
2.2k
Tell your own story through comics
letsgokoyo
1
950
Transcript
&DIP4UBUF/FUXPSLT&YQMJDJU.FNPSZ IJNLU!य़ΤϦΞ DEEP LEARNING BOOK 4FRVFODF.PEFMJOH3FDVSSFOUBOE3FDVSTJWF/FUT
&DIP4UBUF/FUXPSLT w 3//ʹֶ͓͍ͯश͕େมͳύϥϝʔλ w ӅΕӅΕ SFDVSSFOUXFJHIUT w ೖྗӅΕ JOQVUXFJHIUT
w &DIP4UBUF/FUXPSL w ӅΕӅΕॏΈΛݻఆ w ֶश͢Δͷʜ w ೖྗӅΕ JOQVUXFJHIUT w ӅΕग़ྗ PVUQVUXFJHIUT
&DIP4UBUF/FUXPSLT IUUQXXXTDIPMBSQFEJBPSHBSUJDMF&DIP@TUBUF@OFUXPSL
,FSOFMNBDIJOFͱͷྨࣅੑ w Χʔωϧ͕ͬͯΔ͜ͱͬͯʁ w ҙͷ͞ͷܥྻΛݻఆͷϕΫτϧࣸ͢ w ݻఆͷϕΫτϧΛ༻͍ͯྨث͕Λղ͘ w ͜ͷܗͷ߹ɼֶशͷج४ͷઃܭ͕༰қͰ͋Δ w
ग़ྗઢܗճؼͷ߹.4&ͰֶशͰ͖Δ w &4/TೖྗΛԿΒ͔ͷϕΫτϧʹࣸ͢ૢ࡞Λ͍ͯ͠Δ w தͷॏΈݻఆ͍ͯ͠Δ ͍͔ʹաڈͷใΛ๛ʹؚΉදݱ͕ಘΒΕΔ ॏΈΛઃఆ͢ΕΑ͍͔ʁ શવҙຯ͕Θ͔Βͣʜ 3//ΛಈతγεςϜͱΈͳ͢ γεςϜ͕҆ఆ͢ΔΑ͏ͳॏΈΛઃఆ͢Δ
-FBLZ6OJUTBOE0UIFS4USBUFHJFTGPS.VMUJQMF5JNF4DBMF w աڈͷใΛ͑ΔͨΊͷ "EEJOH4LJQ$POOFDUJPOTUISPVHI5JNF w ޯͷফࣦͷ͕͘ͳΔ w രൃݩͷ3//ͱಉ͡Ͱൃੜ͢Δ
-FBLZ6OJUTBOEB4QFDUSVNPG%J⒎FSFOU5JNF4DBMFT w աڈͷใΛͲͷఔ͔͢Λ੍ޚ͢Δ 3FNPWJOH$POOFDUJPOT w ͍࣌ࠁͰͷґଘΛ͍࣌ࠁͰͷґଘʹஔ͖͑Δ
-FBLZ6OJUT w աڈͷใΛͲͷ͘Β͍͔͢Λௐ͢Δ w ҠಈฏۉͷΑ͏ͳ;Δ·͍Λ͢Δ w Ћ͕େ͖͍ ʹ͍ۙ աڈͷใΛΑΓอଘ͢Δ w
Ћ͕খ͍͞ ʹ͍ۙ աڈͷใΛ͙͢ʹࣺͯΔ w Ћదʹܾఆ͢ΔϋΠύʔύϥϝʔλ µ(t) ↵µ(t 1) + (1 ↵)v(t)
-POH4IPSU5FSN.FNPSZ w ࣗݾϧʔϓΛಋೖ͢Δ͜ͱͰޯ͕ফ͑ʹ͘͘͢Δ IUUQDPMBIHJUIVCJPQPTUT6OEFSTUBOEJOH-45.T 3// -45.
(BUFE3FDVSSFOU6OJUT w ٙ-45.ෳࡶ͗͢ΔͷͰͳ͍͔ʁ w (36-45.ΑΓߴɾ-45.ͱಉͷੑೳ w ͲͪΒ͕ྑ͍͔λεΫʹΑΔ -45. (36
IUUQTJTBBDDIBOHIBVHJUIVCJP-45.BOE(36'PSNVMB4VNNBSZ
ࣜతʹൺֱ͢ΔʢόΠΞεΛແࢹʣ -45. (36 zt = (xtUz + ht 1Wz)
rt = (xtUr + ht 1Wr) ˜ ht = tanh ⇣ xt + Uh + (rt ht 1)Wh ⌘ ht = (1 zt) ht 1 + zt ˜ ht it = (xtUi + ht 1Wi) ft = (xtUf + ht 1Wf ) ot = (xtUo + ht 1Wg) ˜ Ct = tanh (xtUg + ht 1Wg) Ct = (ft Ct 1 + it ˜ Ct) ht = tanh (Ct) ot (36Ͱೖྗήʔτͱ٫ήʔτ͕౷߹͞Ε͍ͯΔ
0QUJNJ[BUJPOGPS-POH5FSN%FQFOEFODJFT w 3//Λϕʔεͱͨ͠χϡʔϥϧωοτϫʔΫͷඍ w ඇৗʹେ͖ͳΛͱΔPS w ඇৗʹখ͞ͳΛͱΔ w ಛʹɼޯ͕ඇৗʹେ͖ͳͱ͖ʹͲ͏͢Εྑ͍͔ʁ
ޯͷΫϦοϐϯά ޯͷਖ਼نԽ
$MJQQJOH(SBEJFOU w ޯ͕ඇৗʹ େ͖͍cখ͍͞ ͱʁ w ͍͍ͩͨฏΒ͚ͩͲͱ͖Ͳ͖֑͕͋Δ IUUQXXXEFFQMFBSOJOHCPPLPSHMFDUVSF@TMJEFTIUNM
$MJQQJOH(SBEJFOU w ޯ๏ϕʔεͷख๏ʹΑΔͱʜ w ֑ͷपΓͰ͕ਧ͖ඈΜͰ͠·͏ ޯരൃ w ޯ͕େ͖͘ͳΓ͗ͨ͢ΒޯͷϊϧϜͰׂΔ w
ޯΛHͱͯ͠ʜ w WϋΠύʔύϥϝʔλ ࣗવݴޠॲཧͩͱ͕ଟ͍ g ( gv ||g|| (||g|| > v) g (otherwise)
3FHVMBSJ[JOHUP&ODPVSBHF*OGPSNBUJPO'MPX w ਖ਼ଇԽ߲Λಋೖ͢Δ͜ͱͰʮJOGPSNBUJPOqPXʯΛଅਐ w ͜ͷ߲ͷܭࢉ͍͕͠ɼۙࣅ͕ఏҊ͞Ε͍ͯΔ w $MJQQJOHͱΈ߹ΘͤΔ͜ͱͰهԱͰ͖Δڑ͕৳ͼΔ ⌦ =
X t ⇣||(rh(t) L) @h(t) @h(t 1) || ||(rh(t) L)|| 1 ⌘2
&YQMJDJU.FNPSZ w χϡʔϥϧωοτϫʔΫʜ w ҉తͳใͷอ࣋ಘҙ w ໌ࣔతͳใ ࣄ࣮ ͷอ࣋ۤख w
໌ࣔతͳใΛอ࣋͠ɼਪʹ׆༻͢Δߏ ʢϫʔΩϯάϝϞϦͷಋೖʣ w .FNPSZ/FUXPSLT w /FVSBM5VSJOH.BDIJOF
"TDIFNBUJDPGBOFUXPSLXJUIBOFYQMJDJUNFNPSZ IUUQXXXEFFQMFBSOJOHCPPLPSHMFDUVSF@TMJEFTIUNM
"TDIFNBUJDPGBOFUXPSLXJUIBOFYQMJDJUNFNPSZ w ਖ਼֬ͳϝϞϦͷΞυϨεΛग़ྗ͢Δͷ͍͠ w ଟ͘ͷϝϞϦηϧͷॏΈ͖ฏۉΛͱΔ w ॏΈιϑτϚοΫεͳͲͰ࡞Δ ʢͰ͖Δ͚ͩҰՕॴͷϝϞϦΛࢀর͢ΔΑ͏ʹʣ w ϝϞϦηϧεΧϥΑΓϕΫτϧͷํ͕ྑ͍
w ίϯςϯπϕʔεΞυϨογϯά͕ՄೳʹͳΔ w ʮl8FBMMMJWFJOBZFMMPXTVCNBSJOFzΛؚΉՎࢺΛݟ͚ͭΔʯ w ʢϩέʔγϣϯϕʔεΞυϨογϯάͱʁʣ w ʮεϩοτ347ʹ֨ೲ͞Ε͍ͯΔՎࢺΛऔಘ͢Δʯ w ʢΞυϨογϯάΞςϯγϣϯͱಉ͡ܗࣜʣ