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
2019_G検定対策_数学講座01_ディープラーニングでの基礎計算問題/20190125_JD...
Search
Sponsored
·
Ship Features Fearlessly
Turn features on and off without deploys. Used by thousands of Ruby developers.
→
ITO Akihiro
January 25, 2019
Technology
17
0
Share
2019_G検定対策_数学講座01_ディープラーニングでの基礎計算問題/20190125_JDLA_G_Math_1
G検定対策社内数学講座
--
ディープラーニングでの基礎計算問題
パーセプトロンとCNNと活性化関数
ITO Akihiro
January 25, 2019
More Decks by ITO Akihiro
See All by ITO Akihiro
kintone + ローカルLLM = ?
akit37
0
200
【NoMapsTECH 2025】AI Edge Computing Workshop
akit37
0
680
【NoMapsTECH 2025】AI Tech Community Talk
akit37
0
320
エンジニア目線でのテスラ
akit37
0
79
「重鎮問題」について(軽めに)
akit37
0
84
Software + Hardware = Fun++
akit37
0
64
基本的に "リモートしかない" ワーク/20231128_KBS_LT
akit37
1
45
3つの先端技術が コミュニティ軸で融合した話。/20230615_CMCMeetup
akit37
0
45
Bootleg_越境してみたときのアウェイ感。/20230328_CMCMeetup
akit37
0
52
Other Decks in Technology
See All in Technology
Amazon CloudFrontにおけるAIボットアクセス制御のポイント
kizawa2020
4
290
Claude Codeですべての日常業務を爆速化しよう!
minorun365
PRO
16
14k
管理アカウント単一運用からAWS Organizationsに移行するの大変で滅
hiramax
0
270
DI コンテナ自動生成ツールを実装してみた / intro-autodi
uhzz
0
870
オンコールの負荷軽減のためのBits Assistant 活用方法 / How to Use Bits Assistant to Reduce the Workload on On-Call Staff
sms_tech
1
240
自作エディターをOSSにして分かった、一人に刺さる開発が世界を動かす理由
shinyasaita
1
440
Javaで学ぶSOLID原則
negima
1
200
Copilot CLI・IDE・Web・スマホで途切れない開発フローを目指して / One Copilot flow - CLI IDE Web Mobile
aeonpeople
1
1.1k
OpenClawとHermesAgentでAI新入社員を作った話
takanoriyanada
0
130
人が担う「価値」とは?これからの「QA」とは / Human Value and the Future of Quality Assurance
bitkey
PRO
0
110
Agentic Design Patterns
glaforge
0
230
Javaコミュニティをもっと楽しむための9箇条
takasyou
0
290
Featured
See All Featured
AI in Enterprises - Java and Open Source to the Rescue
ivargrimstad
0
1.3k
The Spectacular Lies of Maps
axbom
PRO
1
770
Max Prin - Stacking Signals: How International SEO Comes Together (And Falls Apart)
techseoconnect
PRO
0
170
Color Theory Basics | Prateek | Gurzu
gurzu
0
320
Public Speaking Without Barfing On Your Shoes - THAT 2023
reverentgeek
1
400
JAMstack: Web Apps at Ludicrous Speed - All Things Open 2022
reverentgeek
1
450
Neural Spatial Audio Processing for Sound Field Analysis and Control
skoyamalab
0
310
Site-Speed That Sticks
csswizardry
13
1.2k
SERP Conf. Vienna - Web Accessibility: Optimizing for Inclusivity and SEO
sarafernandez
2
1.5k
Cheating the UX When There Is Nothing More to Optimize - PixelPioneers
stephaniewalter
287
14k
Building Flexible Design Systems
yeseniaperezcruz
330
40k
Making the Leap to Tech Lead
cromwellryan
135
9.8k
Transcript
ディープラーニングでの 基礎計算 〜パーセプトロンとCNNと活性化関数〜 Jun. 2019 created by ITO Akihiro
パーセプトロンの概念 入力 重み バイアス 出力 入力:1 or 0 出力:1 or 0
画像認識の概念 1 2 3 4 5 6 7 8 9
10 11 12 31 32 12 21 22 23 3 1 2 ・ ・ ・ 入力層 中間層 出力層 1 0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 0 0 重み w1-1 ... w1-36 重み w2-1 ... w2-6 それぞれのシグモイドニューロンは、 ある特徴に対して発火する
CNN (Convolutional Neural Network)の概念 畳み込み層 フィル タ3 特徴マップ 3 プーリング層
出力層 ユニット3 特徴を抽出 特徴をさらに濃縮
フィルタを順に当て 画像を単純化し、 特徴を抽出する。 1 2 3 4 5 6 7
8 9 10 11 12 13 14 15 16 17 18 19 20 21 ... ... 121 元画像 フィルタ1 フィルタ2 斜め方向に反応 縦方向に反応
例: 入力画像 (H,W)=(11,11) フィルタサイズ(Fh,Fw)=(3,3) パディング (P)=0 ストライド (S)=2 ⇒ 出力画像 (Oh,Ow)=(5,5) 1 0 0
1 0 1 3 1 2 1 2 2 3 2 3 2 2 2 3 2 1 2 0 1 1 0 0 0 0 0 0 2 1 2 0 2 2 3 2 2 1 2 2 2 1 0 1 0 1 0 公式
活性化関数 ステップ シグモイド tanh (ハイパボリックタンジェント) ReLU ReLU6 0 1 0
0 0 0 6 0 0 1 0 -1 1 0