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
ITO Akihiro
January 25, 2019
Technology
0
4
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
Tweet
Share
More Decks by ITO Akihiro
See All by ITO Akihiro
「重鎮問題」について(軽めに)
akit37
0
45
Software + Hardware = Fun++
akit37
0
22
基本的に "リモートしかない" ワーク/20231128_KBS_LT
akit37
1
18
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
15
Bootleg_越境してみたときのアウェイ感。/20230328_CMCMeetup
akit37
0
21
始まりは2017年のG検定。/20221026_AITable
akit37
0
13
kintone知能化計画/20220902_kintone_and_JPStripes
akit37
0
20
外観検査用画像前処理の_コツをコード解説付きで。/20220810_CDLE_LT
akit37
0
15
サブスク課金に銀行振込を追加してみた。その①/20220713_JPStripes
akit37
0
19
Other Decks in Technology
See All in Technology
RayでPHPのデバッグをちょっと快適にする
muno92
PRO
0
200
株式会社Awarefy(アウェアファイ)会社説明資料 / Awarefy-Company-Deck
awarefy
3
12k
Охота на косуль у древних
ashapiro
0
130
いまからでも遅くない!コンテナでWebアプリを動かしてみよう!コンテナハンズオン編
nomu
0
180
開発者のための FinOps/FinOps for Engineers
oracle4engineer
PRO
2
260
OPENLOGI Company Profile for engineer
hr01
1
20k
エンジニアの健康管理術 / Engineer Health Management Techniques
y_sone
4
730
エンジニアのキャリアパスと、 その中で自分が大切にしていること
noteinc
3
230
【Forkwell】「正しく」失敗できるチームを作る──現場のリーダーのための恐怖と不安を乗り越える技術 - FL#83 / A team that can fail correctly by forkwell
i35_267
2
110
Global Databaseで実現するマルチリージョン自動切替とBlue/Greenデプロイ
j2yano
0
170
IAMのマニアックな話2025
nrinetcom
PRO
6
1.4k
MLflowはどのようにLLMOpsの課題を解決するのか
taka_aki
0
140
Featured
See All Featured
BBQ
matthewcrist
87
9.5k
The World Runs on Bad Software
bkeepers
PRO
67
11k
Sharpening the Axe: The Primacy of Toolmaking
bcantrill
40
2k
Why You Should Never Use an ORM
jnunemaker
PRO
55
9.2k
Making Projects Easy
brettharned
116
6k
ピンチをチャンスに:未来をつくるプロダクトロードマップ #pmconf2020
aki_iinuma
115
51k
VelocityConf: Rendering Performance Case Studies
addyosmani
328
24k
Git: the NoSQL Database
bkeepers
PRO
428
65k
Distributed Sagas: A Protocol for Coordinating Microservices
caitiem20
330
21k
StorybookのUI Testing Handbookを読んだ
zakiyama
28
5.5k
Navigating Team Friction
lara
183
15k
Designing on Purpose - Digital PM Summit 2013
jponch
117
7.1k
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