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
About Spectral Clustering
Search
Shunya Ueta
October 16, 2014
Research
2.9k
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
About Spectral Clustering
Spectral Clusteringというクラスタリング手法についての基本的な説明のスライドです。
@MMA_LAB
Shunya Ueta
October 16, 2014
More Decks by Shunya Ueta
See All by Shunya Ueta
20分で分かる Human-in-the-Loop 機械学習におけるアノテーションとヒューマンコンピューターインタラクションの真髄
hurutoriya
6
4k
Query Understanding for Search Engines. Chap2 Query Classification
hurutoriya
0
590
Introducing "Challenges and research opportunities in eCommerce search and recommendations"
hurutoriya
0
360
Auto Content Moderation in C2C e-Commerce at OpML20
hurutoriya
0
780
TFX: A tensor flow-based production-scale machine learning platform
hurutoriya
0
690
Applied machine learning at facebook a datacenter infrastructure perspective HPCA18
hurutoriya
0
350
machine learning tips in the python world PRMLer Night
hurutoriya
1
720
パターン認識と機械学習 第1章 #PRML学ぼう PRML輪講 #2 / PRML Seminar 2 go to introduction in machine learning
hurutoriya
1
3.2k
複数人でコードを書く際のFist Step
hurutoriya
0
420
Other Decks in Research
See All in Research
研究室単位での自律的 IPv6接続性確立に向けたAS共同運用モデルの提案と実証
reokashiwa
PRO
0
210
RS-Agent: Automating Remote Sensing Tasks through Intelligent Agent
satai
3
530
Model Discovery and Graph Simulation: A Lightweight Gateway to Chaos Engineering
anatolykr
0
300
CVPR2026論文紹介_VLMにとって良いvision encoderとは何か?Rethinking Model Selection in VLM Through the Lens of Gromov-Wasserstein Distance
kobayashi31
1
210
マーケットストリート 社会実験2024 in 秋葉原ジャンク通り 調査報告書
izumiyama_lab
1
140
HAKARI-Bench - 実運用視点での情報検索モデル評価ベンチマーク
hotchpotch
1
730
Data Visualization Tools in the Age of AI
flekschas
0
200
【Zozo Research 技術共有会】三次元領域の現在と展望
mickey_0226
3
600
J-STAGEの現況と全文XML登載必須化について
xspa2012
0
200
【ローカルAIに向き合う展示会vol.2】液体時間定数型モジュールを用いた オリジナルの双方向エンコーダーモデルNexteraBERT 推論速度向上検討並びにダウンストリーム評価
rikkabotan7
0
190
Spatial Active Noise Control Based onSound Field Interpolation Incorporating Physical Constraints
skoyamalab
0
180
超効率化への挑戦:1bit LLMの現状と展望
yumaichikawa
0
680
Featured
See All Featured
Producing Creativity
orderedlist
PRO
348
41k
A Tale of Four Properties
chriscoyier
163
24k
Building Better People: How to give real-time feedback that sticks.
wjessup
370
20k
Fantastic passwords and where to find them - at NoRuKo
philnash
52
3.8k
HDC tutorial
michielstock
2
850
Helping Users Find Their Own Way: Creating Modern Search Experiences
danielanewman
31
3.3k
Music & Morning Musume
bryan
47
7.4k
Marketing to machines
jonoalderson
1
5.7k
The Illustrated Children's Guide to Kubernetes
chrisshort
51
53k
Tell your own story through comics
letsgokoyo
1
1.1k
Why Mistakes Are the Best Teachers: Turning Failure into a Pathway for Growth
auna
0
250
Agile Actions for Facilitating Distributed Teams - ADO2019
mkilby
0
270
Transcript
About Spectral Clustering Univ. of Tsukuba MMA Lab Shunya
Ueta
目次 1. Spectral Graph 1. About 2. Graph
Laplacian Matrix 3. 応用例 2. Spectral Clustering 3. 実装 2
About Spectral Graph 歴史: 1950年代~ 目的:
グラフの特徴とグラフの固有値・固有ベクトル を結びつける 応用例: Spectral Clustering 画像領域分割 3
Graph Laplacian matrix 4 定義: 4 2 3
5 1 [ 0 1 0 1 0 ] [ 1 0 1 1 0 ] [ 0 1 0 0 1 ] [ 1 1 0 0 1 ] [ 0 0 1 1 0 ] [ 2 0 0 0 0 ] [ 0 3 0 0 0 ] [ 0 0 2 0 0 ] [ 0 0 0 3 0 ] [ 0 0 0 0 2 ] AG DG G n頂点無向グラフ G = (V, E) : AG : DG : Gの近接行列 Gの次数行列 LG = DG AG ラプラシアン行列 [ 2 -‐1 0 -‐1 0 ] [ -‐1 3 -‐1 -‐1 0 ] [ 0 -‐1 2 0 -‐1 ] [ -‐1 -‐1 0 3 -‐1 ] [ 0 0 -‐1 -‐1 2 ] LG = =
Spectral Graphの応用例 画像領域分割 : 画素毎の類似画像 5
Spectral Graphの応用例 画像領域分割 6
Spectral Grapth の応用例 Spectral Clustering 7
Spectral Clustering 8 P次元 n 個 .
. . 目的 p次元のデータn個を kクラスタに分類したい グラフ表現 データをグラフで表す Laplacian matrix グラフの行列表現 n n 対称行列 固有値集合(スペクトラム)を求める 固有ベクトルを小さいものから k番目までを選定 k n
Graph Laplacian matrix 9 定義: 4 2 3
5 1 [ 0 1 0 1 0 ] [ 1 0 1 1 0 ] [ 0 1 0 0 1 ] [ 1 1 0 0 1 ] [ 0 0 1 1 0 ] [ 2 0 0 0 0 ] [ 0 3 0 0 0 ] [ 0 0 2 0 0 ] [ 0 0 0 3 0 ] [ 0 0 0 0 2 ] AG DG G n頂点無向グラフ G = (V, E) : AG : DG : Gの近接行列 Gの次数行列 LG = DG AG ラプラシアン行列 [ 2 -‐1 0 -‐1 0 ] [ -‐1 3 -‐1 -‐1 0 ] [ 0 -‐1 2 0 -‐1 ] [ -‐1 -‐1 0 3 -‐1 ] [ 0 0 -‐1 -‐1 2 ] LG = =
ProperLes of Laplacian matrix 10 1. Lは部分対角優位行列なので全ての固有値は 0 以上
2. L の最小固有値は 0 であり、対応する固有ベクトルは全要素1の n 次元ベクトル 3. 固有値0の個数はグラフの連結部の数 連結部A 連結部B Laplacian matrix グラフの行列表現 対称行列 A B 1 1 1 0 0 0 I 0 = x x 固有値 対角行列
ProperLes of Laplacian matrix 11 1. Lは部分対角優位行列なので全ての固有値は 0 以上
2. L の最小固有値は 0 であり、対応する固有ベクトルは全要素1の n 次元ベクトル 3. 固有値0の個数はグラフの連結部の数 連結部A 連結部B Laplacian matrix グラフの行列表現 対称行列 A B 0 0 0 1 1 1 I 0 = x x 固有値 対角行列
Spectral Clustering 12 理想的なグラフ状態 各行に対して各列の要素が クラスタを示している
k n 1 1 0 0 0 0 0 1 0 0 i番目の行はあるデータX_i が所属するクラスタを 示している 0 0 0 1 1