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
Rで有名絵画を安全に買いたい
Search
saltcooky
September 16, 2022
Science
0
280
Rで有名絵画を安全に買いたい
TokyoR #101 LT
saltcooky
September 16, 2022
Tweet
Share
More Decks by saltcooky
See All by saltcooky
FIBA W杯の日本代表って組み合わせ次第で2次ラウンド行けたんじゃね?をデータで検証
saltcooky12
0
260
階層クラスタリングにおける仮説検定
saltcooky12
0
920
データドリブンな仮説検証のためのSelective Inference
saltcooky12
1
1.2k
ストリートスナップデータに 統計的ネットワーク分析の適用を試みた
saltcooky12
0
770
Other Decks in Science
See All in Science
[第62回 CV勉強会@関東] Long-CLIP: Unlocking the Long-Text Capability of CLIP / kantoCV 62th ECCV 2024
lychee1223
1
670
Snowflakeによる統合バイオインフォマティクス
ktatsuya
0
490
Transformers are Universal in Context Learners
gpeyre
0
550
(Forkwell Library #48)『詳解 インシデントレスポンス』で学び倒すブルーチーム技術
scientia
2
1.4k
重複排除・高速バックアップ・ランサムウェア対策 三拍子そろったExaGrid × Veeam連携セミナー
climbteam
0
110
ICRA2024 速報
rpc
3
5.2k
【人工衛星】座標変換についての説明
02hattori11sat03
0
110
生成AI による論文執筆サポートの手引き(ワークショップ) / A guide to supporting dissertation writing with generative AI (workshop)
ks91
PRO
0
250
Direct Preference Optimization
zchenry
0
280
All-in-One Bioinformatics Platform Realized with Snowflake ~ From In Silico Drug Discovery, Disease Variant Analysis, to Single-Cell RNA-seq
ktatsuya
0
230
(2024) Livres, Femmes et Math
mansuy
0
110
Causal discovery based on non-Gaussianity and nonlinearity
sshimizu2006
0
190
Featured
See All Featured
[RailsConf 2023 Opening Keynote] The Magic of Rails
eileencodes
28
9.1k
Building Applications with DynamoDB
mza
90
6.1k
Fireside Chat
paigeccino
34
3k
Bootstrapping a Software Product
garrettdimon
PRO
305
110k
The Myth of the Modular Monolith - Day 2 Keynote - Rails World 2024
eileencodes
16
2.1k
Distributed Sagas: A Protocol for Coordinating Microservices
caitiem20
329
21k
Embracing the Ebb and Flow
colly
84
4.5k
A Tale of Four Properties
chriscoyier
156
23k
Learning to Love Humans: Emotional Interface Design
aarron
273
40k
Designing on Purpose - Digital PM Summit 2013
jponch
115
7k
4 Signs Your Business is Dying
shpigford
180
21k
Product Roadmaps are Hard
iamctodd
PRO
49
11k
Transcript
3Ͱ༗໊ֆըΛ҆શʹങ͍͍ͨ !TBMUDPPLZ 5PLZP3 1
୭ʁ 2 !TBMUDPPLZ • 3ྺɿ͙Β͍͔ͳ • ۈઌɿຊʹ͋Δ*5ܥͷձࣾ • ࣄ༰ɿ3%తͳ෦ॺͰ
ɹɹɹ3Λͬͨσʔλੳ͞Μ ػցֶशͷॲཧ࡞ • झຯɿϑΝογϣϯඒज़ؗ८Γ
δϟΫιϯϙϩοΫΛ͍ͬͯ·͔͢ 3 +BDLTPO1PMMPL நදݱओٛͷදతͳΞϝϦΧਓըՈ
δϟΫιϯϙϩοΫΛ͍ͬͯ·͔͢ 4 υϩοϓϖΠϯςΟϯά
δϟΫιϯϙϩοΫΛ͍ͬͯ·͔͢ 5 ʰ/P ʱ ºNN
δϟΫιϯϙϩοΫΛ͍ͬͯ·͔͢ 6 ʰ/P ʱ ºNN ݄ ݱඒज़࠷ߴֹ ࣌ ԯສυϧ
δϟΫιϯϙϩοΫΛ͍ͬͯ·͔͢ 7 ཉ͍͠ʂ
δϟΫιϯϙϩοΫΛ͍ͬͯ·͔͢ 8 ͚Ͳɺِଟͦ͏ʜ
ϙϩοΫͷֆըΛղੳͨ͠ݚڀ 9 • Fractal analysis of Pollock’s drip paintings.
(R.P.Taylor, et al , 1999) • On multifractal structure in non-representational art. (J.R.Nureika, et al , 2005) ˠϙϩοΫͷυϩοϓϖΠϯςΟϯάʹ ϑϥΫλϧߏ͕͋Δ͜ͱ͕Θ͔Δ
ϑϥΫλϧߏ 10 ਤܗͷҰ෦Λ֦େ͢Δͱɺશମͱ૬ࣅ͢Δܗ ࣗݾ૬ࣅੑ ͕ଘࡏ͢Δߏ FYγΣϧϐϯεΩʔͷΪϟεέοτ
ϑϥΫλϧ࣍ݩ ༰ྔ࣍ݩϋυϧϑ࣍ݩ 11 w ͲΕ͚ͩࣗݾ૬ࣅੑ͕͋Δ͔Λࣔ͢ྔ w ֤ۭؒํʹ-ʹॖΊΔͱɺͱͷਤܗΛຒΊΔʹ/-%ݸ ͷࣗݾ૬ࣅਤܗ͕ඞཁͱ͍͏͜ͱΛදݱ w
ʙͷؒΛͱΓɺʹ͍ۙ΄Ͳࣗݾ૬ؔੑ͕ڧ͍ w γΣϧϐϯεΩʔͷΪϟεέοτͷ࣍ݩ w ղੳతʹ#PY$PVOUJOHΞϧΰϦζϜʹΑΓਪఆ
#PY$PVOUJOHΞϧΰϦζϜ 12 ϑϥΫλϧ࣍ݩΛٻΊΔྲྀΕ ̍ରը૾ΛҰลͷ͕͞-ͷϒϩοΫʹ͚Δ
#PY$PVOUJOHΞϧΰϦζϜ 13 ϑϥΫλϧ࣍ݩΛٻΊΔྲྀΕ ̍ରը૾ΛҰลͷ͕͞-ͷϒϩοΫʹ͚Δ
#PY$PVOUJOHΞϧΰϦζϜ 14 ϑϥΫλϧ࣍ݩΛٻΊΔྲྀΕ ̎ର͕өΓࠐΜͰ͍ΔϒϩοΫͷ/ - Λ͑Δ
#PY$PVOUJOHΞϧΰϦζϜ 15 ϑϥΫλϧ࣍ݩΛٻΊΔྲྀΕ ϒϩοΫେ͖͞-Λখͯ͘͞͠ରը૾Λ͚Δ
#PY$PVOUJOHΞϧΰϦζϜ 16 ϑϥΫλϧ࣍ݩΛٻΊΔྲྀΕ ର͕өΓࠐΜͰ͍ΔϒϩοΫͷ/ - Λ͑Δ
#PY$PVOUJOHΞϧΰϦζϜ 17 ϑϥΫλϧ࣍ݩΛٻΊΔྲྀΕ େ͖͞-Λখ͘͞͠ͳ͕Βର͕өΔϒϩοΫΛΧϯτ͢Δ ɹ͜ͱΛ܁Γฦ͢
#PY$PVOUJOHΞϧΰϦζϜ 18 ϑϥΫλϧ࣍ݩΛٻΊΔྲྀΕ ϒϩοΫͷେ͖͞-ͱΧϯτ/ - ͷ྆ରάϥϑʹ͓͚Δ ɹճؼઢͷ͖͕ϑϥΫλϧ࣍ݩʹͳΔ log N(L)
= D log(L) + K ʢ,ఆʣ MPH/ - MPH- ޯ%ʹ ϑϥΫλϧ࣍ݩ
19 w ͳͥ͜ΕͰϑϥΫλϧ࣍ݩΛਪఆ͢Δ͜ͱ͕Ͱ͖Δͷ͔ ఆ͔ٛΒͷมܗ log N(L) = log( a
L )D log N(L) = D log(L) + D log(a) #PY$PVOUJOHΞϧΰϦζϜ ʢBɿਖ਼ͷఆʣ
20 δϟΫιϯϙϩοΫͷ࡞ͷ߹ w -DNͰϑϥΫλϧ࣍ݩ͕มԽ w %% -Ҏ্ %- -ະຬ ʰ#MVF1PMFT/VNCFS
ʱ MPH - NN MPH / #PY$PVOUJOHΞϧΰϦζϜ
ϙϩοΫͷֆըͷಛ 21 ʢ̍ʣ̎छͷϑϥΫλϧύλʔϯ͔ΒΔ ʢ̎ʣ༷ʑͳεέʔϧʹ͓͍ͯϑϥΫλϧੑ͕ଘࡏ ʢ̏ʣϑϥΫλϧ࣍ݩରάϥϑͷޯ͔ΒٻΊΕΔ ʢ̐ʣ%-ʼ%% ʢ̑ʣۙࣅۂઢͷඪ४ภ͕ࠩখ͍͞ d ʢ̒ʣ֤৭ͷͰ্هͷ̑ͭͷಛΛຬͨ͢
3Ͱ#PY$PVOUJOH 22 7PY31BDLBHF
3Ͱ#PY$PVOUJOH 23 γΣϧϐϯεΩʔͷΪϟεέοτͷϑϥΫλϧ࣍ݩΛٻΊΔ ը૾ॲཧ
3Ͱ#PY$PVOUJOH 24 γΣϧϐϯεΩʔͷΪϟεέοτͷϑϥΫλϧ࣍ݩΛٻΊΔ ϑϥΫλϧ࣍ݩΛٻΊΔ
3Ͱ#PY$PVOUJOH 25 γΣϧϐϯεΩʔͷΪϟεέοτͷϑϥΫλϧ࣍ݩΛٻΊΔ ྆ରάϥϑͱճؼઢͷՄࢹԽ 0 2 4 6 8 -7
-6 -5 -4 -3 -2 log(1/res) log(N) Box Counting method : D=1.5747
؆୯ʹ·ͱΊ 26 w δϟΫιϯϙϩοΫͷֆըʹϑϥΫλϧߏ͕ଘࡏ͢Δ w ϑϥΫλϧߏͷϑϥΫλϧ࣍ݩΛղੳతʹٻΊΔͨΊʹ ɺ#PY$PVOUJOHΞϧΰϦζϜΛ༻͍Δ w
3Ͱ7PY3QBDLBHFͷCPY@DPVOUJOHؔͰ࣮ߦͰ͖Δ
&/% 27 &OKPZ
ࢀߟࢿྉ 28 •RʹΑΔը૾ॲཧɿimagerύοέʔδͷ͍ํ https://htsuda.net/archives/1985 •ϘοΫεΧϯτ๏ʹΑΔඐഀބͷϑϥΫλϧ࣍ݩ https://shiga-u.repo.nii.ac.jp/?action=repository_uri&item_id=1751