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
himkt
January 29, 2018
Research
2
180
Deep Learning Book 10その2 / deep learning book 10 vol2
himkt
January 29, 2018
Tweet
Share
More Decks by himkt
See All by himkt
Linformer: paper reading
himkt
0
440
RoBERTa: paper reading
himkt
1
330
NLP SoTA 勉強会 / ner_2019
himkt
2
1.4k
自然言語処理 @ クックパッド / nlp at cookpad
himkt
1
500
Interpretable Machine Learning 6.3 - Prototypes and Criticisms
himkt
2
150
ニューラル固有表現抽出 / Neural Named Entity Recognition
himkt
3
690
ニューラル固有表現抽出器を実装してみる / PyNER
himkt
6
2.1k
Spacyでお手軽NLP / NLP with spacy
himkt
0
1k
ふわふわ系列ラベリング / ner 2018
himkt
5
850
Other Decks in Research
See All in Research
Gemini と Looker で営業DX をドライブする / Driving Sales DX with Gemini and Looker
sansan_randd
0
160
セミコン地域における総合交通戦略
trafficbrain
0
130
Intrinsic Self-Supervision for Data Quality Audits
fabiangroeger
0
390
書き手はどこを訪れたか? - 言語モデルで訪問行動を読み取る -
hiroki13
0
160
20250226 NLP colloquium: "SoftMatcha: 10億単語規模コーパス検索のための柔らかくも高速なパターンマッチャー"
de9uch1
0
170
言語モデルによるAI創薬の進展 / Advancements in AI-Driven Drug Discovery Using Language Models
tsurubee
1
140
Segment Any Change
satai
3
260
メタヒューリスティクスに基づく汎用線形整数計画ソルバーの開発
snowberryfield
3
800
言語モデルLUKEを経済の知識に特化させたモデル「UBKE-LUKE」について
petter0201
0
270
BtoB プロダクトにおけるインサイトマネジメントの必要性 現場ドリブンなカミナシがインサイトマネジメントに取り組むワケ / Why field-driven Kaminashi is working on insight management
kaminashi
1
340
ラムダ計算の拡張に基づく 音楽プログラミング言語mimium とそのVMの実装
tomoyanonymous
0
430
[ECCV2024読み会] 衛星画像からの地上画像生成
elith
1
1.1k
Featured
See All Featured
Large-scale JavaScript Application Architecture
addyosmani
511
110k
How GitHub (no longer) Works
holman
314
140k
Art, The Web, and Tiny UX
lynnandtonic
298
20k
Fashionably flexible responsive web design (full day workshop)
malarkey
406
66k
How to train your dragon (web standard)
notwaldorf
91
5.9k
For a Future-Friendly Web
brad_frost
176
9.6k
Code Review Best Practice
trishagee
67
18k
The World Runs on Bad Software
bkeepers
PRO
67
11k
Navigating Team Friction
lara
183
15k
VelocityConf: Rendering Performance Case Studies
addyosmani
328
24k
The Web Performance Landscape in 2024 [PerfNow 2024]
tammyeverts
4
470
How STYLIGHT went responsive
nonsquared
99
5.4k
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