Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Sign up for free
Menu
Search
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Pricing
Search
Sign in
Sign up for free
190306_2章_ver2
Search
Sponsored
·
Your Podcast. Everywhere. Effortlessly.
Share. Educate. Inspire. Entertain. You do you. We'll handle the rest.
→
makoto
March 06, 2019
34
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
190306_2章_ver2
makoto
March 06, 2019
More Decks by makoto
See All by makoto
190309_2_2.9-12
makoto8
0
4
190306_2章
makoto8
0
11
Featured
See All Featured
Reflections from 52 weeks, 52 projects
jeffersonlam
356
22k
Dealing with People You Can't Stand - Big Design 2015
cassininazir
367
27k
From π to Pie charts
rasagy
1
400
Statistics for Hackers
jakevdp
799
230k
[RailsConf 2023 Opening Keynote] The Magic of Rails
eileencodes
31
10k
Optimising Largest Contentful Paint
csswizardry
37
4k
Product Roadmaps are Hard
iamctodd
55
13k
How to build an LLM SEO readiness audit: a practical framework
nmsamuel
2
950
DBのスキルで生き残る技術 - AI時代におけるテーブル設計の勘所
soudai
PRO
68
58k
The Anti-SEO Checklist Checklist. Pubcon Cyber Week
ryanjones
0
270
Navigating Team Friction
lara
192
16k
Faster Mobile Websites
deanohume
310
32k
Transcript
2019/03/06 ʮਂֶशʯୈ2ষ 2-1~2-8
εΧϥʔʢscalarsʣ ୯Ұͷ ͲΜͳछྨͷͳͷ͔໌ࣔ ▸ ࣮ͷεΧϥʔ ▸ ࣗવͷεΧϥʔ 2 s ∈
ℝ n ∈ ℕ
ϕΫτϧʢvectorsʣ ͷྻ ▸ ্ۭؒͷΛද͍ͯ͠ΔͱΈͳͤΔ ▸ ͦΕͧΕͷཁૉ֤࣠ͷ࠲ඪ ཁૉͷू߹ʹΠϯσοΫεΛৼΔ߹ ▸ ΠϯσοΫεΛؚΉू߹Λఆٛ ɹɹɹɹɹɹΛରͱ͢Δ࣌ɼू߹ɹɹɹɹɹΛఆٛˠ
▸ ه߸͋Δू߹ͷิू߹ ɹɹɹɹɹͷ͏ͪɹҎ֎ͷશཁૉΛؚΉϕΫτϧ ɹɹɹɹɹͷ͏ͪɹɹɹɹҎ֎ͷશཁૉΛؚΉϕΫτϧ 3 x1 , x3 , x6 S = {1,3,6} x−1 x x1 x−S x x1 , x3 , x6 xS
ߦྻʢmatricesʣ ͭͷΠϯσοΫεͰࣔ͞ΕΔྻ ▸ ࣮ߦྻɹͷߴ͕͞ɹͰ෯͕ ▸ ɿͯ͢ͷཁૉΛࢦఆɹ ▸ ɹɹͷਨ࠲ඪɹʹ͓͚Δਫฏͳஅ໘ʢɹ൪ͷߦʣ 4 A
m n A ∈ ℝm×n A = [ A1,1 A1,2 A2,1 A2,2 ] (m = n = 2) Ai,: A i i
ςϯιϧʢUFOTPSTʣ ͭҎ্ͷ࣠Λ࣋ͭྻ ͭͷ࣠ͷೖྗʹରͯ͠ ग़ྗ͕͍͔ͭ͘ͷ࣠ʹग़Δ͜ͱΛࣜͰද͢ ྫ͑ʹରͯ͋͠Δํ͔ΒྗΛՃ͑Δ ▸ ྗ͕ՃΘͬͨͷํͳͷʹର͠ ͕มܗ͢ΔมԽྔํٻΊΔඞཁ 5
సஔʢtransposeʣ ߦྻʹؔ͢Δԋࢉͷͭ ▸ ߦྻͷର֯ઢʹର͢Δڸ૾ ϕΫτϧͷసஔ ▸ ϕΫτϧྻͷߦྻˠసஔߦͷߦྻ εΧϥʔͷసஔ ▸ ཁૉΛ͚ͭͩ࣋ͭߦྻˠ
6 A = A1,1 A1,2 A2,1 A2,2 A3,1 A3,2 㱺 AT = [ A1,1 A2,1 A3,1 A1,1 A2,2 A3,2 ] (AT)i,j = Aj,i aT = a
ߦྻͷՃࢉ ▸ ରԠ͢ΔཁૉΛ͢ ɹɹɹɹɹɼ ߦྻͷεΧϥʔͷՃࢉɼࢉ ▸ ֤ཁૉʹରͯ͠ԋࢉΛߦ͏ ɹɹɹɹɹɹɼ 7 సஔʢtransposeʣ
C = A + B Ci,j = Ai,j + Bi,j D = a ⋅ B + c Di,j = a ⋅ Bi,j + c ਂֶशͷจ຺Ͱʜ ɹɹɹɹʢϕΫτϧɹΛߦྻͷ֤ߦʹՃࢉʣΛڐ༰ ɹ৽ͨͳߦྻͷఆ͕ٛෆཁ C = A + b b ϒϩʔυΩϟεςΟϯά
ߦྻͱϕΫτϧͷࢉ ߦྻɹͱɹͷੵ͕ߦྻ ▸ ɹͷྻͷɹɹͷߦͷ ▸ ɹ͕ɹɹɹɼɹ͕ɹɹɹͳΒɼɹ ͭͷϕΫτϧɹɹͷυοτੵ ▸ ߦྻͷੵɹɹͰٻ·Δ ▸
ɹɹɹɹͷɹߦͱɹͷɹྻͷੵͱΈͳͤΔ 8 A B C A = B A m × n B n × p C m × p C = AB Ci,j = ∑ k Ai,k Bk,j x, y xTy C = AB A i B j
ߦྻͷϕΫτϧͷࢉ ɹɹɹɹඞͣ͠Γཱͨͳ͍ɹ ▸ ͭͷϕΫτϧؒͷੵՄ ߦྻͷੵͷసஔ 9 AB = BA xTy
= yTx (1) (ATB) = BTA (2) ɹͷεΧϥʔͰͦΕࣗͷసஔʹ͍͠ ɹΑΓ xTy = (xTy)T = yTx (1) (2)
୯Ґߦྻͱٯߦྻ ୯Ґߦྻ ▸ ɹ࣍ݩͷ୯Ґߦྻ ▸ ҙͷϕΫτϧʹֻ͚ͯΛม͑ͳ͍ߦྻ ٯߦྻ ▸ ɹɹͱද͞Εɼఆٛ 10
n In In x = x I3 = [ 1 0 0 0 1 0 0 0 1 ] ओର֯ઢʹԊ͕ͬͨ A−1 A−1A = In
ٯߦྻ๏ ઢܗํఔࣜɹɹɹɹΛٯߦྻΛ༻͍ͯղ͘ 11 Ax = b A−1A = In ΑΓ
A−1Ax = A−1b Ax = b In x = A−1b x = A−1b ɹɹཧతͳखஈͱͯ͠༗༻ σδλϧͳίϯϐϡʔλͰɹɹͷਫ਼͕ݶఆ͞ΕΔ ɹͷΛར༻ A−1 A−1 b
ɹɹɹɹΛٯߦྻΛ༻͍ͯղ͚Δ ˣ ํఔࣜͷղ͚̍ͭͩղΛ࣋ͭ ઢܗ݁߹ ɹɹɹ ▸ ͍͔ͭ͘ͷϕΫτϧ͔ΒͳΔू߹ͷઢܗ݁߹ɹɹɹɹɹɹ 12 Ax =
b ɹͷྻ͕ɼݪ͔ΒͷํΛ͍ࣔͯͯ͠ ɹʹ౸ୡ͢Δܦ࿏ͷΛܾΊΔͷͩͱߟ͑Δ ɹͷ֤ཁૉͦΕͧΕͷํʹਐΉڑ A b x Ax = ∑ i xi A:,j xi ྻɹͷํʹਐΉڑ i ∑ i ci v(i)
ઢܗैଐͱுΔۭؒ ுΔۭؒ ▸ ݩͷϕΫτϧͷઢܗ݁߹ʹΑͬͯಘΒΕΔ ͔ΒͳΔू߹ ▸ ɹɹɹʹղ͕͋Δ͔ɼɹ ɹ͕ɹͷྻ͕షΔۭؒʹؚ·ΕΔ͔Ͳ͏͔ ઢܗैଐʢੑʣ ▸
ྫ͑ºߦྻʹ͓͍ͯɼ ରͱͳΔɹ࣍ݩ͕ͩɼɹ࣍ݩ ɹˠ͕ுΔۭؒͷ࣍ݩฏ໘Ͱ͔͠ಈ͚ͳ͍ ઢܗಠཱ ▸ ͲͷϕΫτϧଞͷϕΫτϧͱઢܗ݁߹Ͱͳ͍ 13 Ax = b A b A b x
ߦྻ͕ٯߦྻΛ࣋ͭ݅ ▸ ਖ਼ํߦྻ͕ਖ਼ଇͰͳ͚ΕͳΒͳ͍ ˠྻϕΫτϧޓ͍ʹઢܗಠཱ ͷྻϕΫτϧΛฒͨϕΫτϧΛɹɹɹɹɹɹɹɹɹͱ͢Δͱ ɹਖ਼ଇߦྻ͔ͩΒ ٯߦྻͷ݅ 14 A A−1A
= AA−1 = I ͜ͷͱ͖ Ax = 0 Λຬͨ͢ x x = 0 ͷΈ A A = [a1 a2 . . . an] x1 a1 + x2 a2 + . . . + xn an = 0 ˠ x1 = x2 = . . . = xn = 0 ΑͬͯϕΫτϧͷɹɹɹɹɹɹઢܗಠཱͰ͋Δ a1 , a2 , . . . , an
ϊϧϜʢnormʣ ϕΫτϧͷେ͖͞ΛٻΊΔؔ ϢʔΫϦουϊϧϜ ▸ ɹɹɹͰ͋ΔɹϊϧϜ ▸ ݪ͔ΒɹͰࣔ͞ΕΔ·ͰͷϢʔΫϦουڑ ▸ ϕΫτϧͷେ͖͞ɹϊϧϜͷ 15
∥x∥p = (∑ i |xi |p )1 p (p ∈ ℝ, p ≥ 1) p = 2 Lp L2 x = L2 = xTx
ɹϊϧϜ ▸ θϩͷཁૉͱθϩҎ֎ͷཁૉͷ۠ผ͕ॏཁ ▸ ཁૉ͕͔ΒΕ͚ͨͩ૿Ճ͢ΔɹϊϧϜ ▸ ϕΫτϧͷඇཁૉͷ ࠷େϊϧϜ ▸ ϕΫτϧͷதͰઈର͕࠷େͷཁૉͷઈର
16 ϊϧϜʢnormʣ L1 ∥x∥1 = ∑ i |xi | L1 L∞ ∥x∥∞ = max i |xi |
ಛผͳߦྻͱϕΫτϧ ର֯ߦྻ ▸ ओର֯ͷΈඇͰ͋Δߦྻ ▸ ୯Ґߦྻͯ͢ͷର͕֯ ▸ ֻ͚ࢉͷܭࢉޮ͕ྑ͍ ରশߦྻ ▸
సஔߦྻ͕ݩͷߦྻͱಉ͡ 17 Ai,j = Aj,i diag(v) ର֯ߦྻɹɹɹɹͱɹΛֻ͚߹ΘͤΔͱ͖ ֤ཁૉɹΛɹഒ͢Δ͚ͩ x xi vi
ಛผͳߦྻͱϕΫτϧ ୯ҐϕΫτϧ ▸ ͞ʢ୯ҐϊϧϜʣ͕ͷϕΫτϧ ਖ਼نަ ▸ ϕΫτϧɹͱϕΫτϧɹ͕ɹɹͷͱ͖ަ ▸ ޓ͍ʹͷ֯Λ͍ͯ͠Δ ▸
୯ҐϊϧϜ࣋ͭͱ͖ɼਖ਼نަ ަߦྻ ▸ ਖ਼ํߦྻ͔ͭਖ਼نަ 18 xTy x y ATA = AAT = I A−1 = AT ˣ
ݻ༗ղ ߦྻղͷͭ ▸ ͷྻ͔Βݟ͑ͳ͍ػೳతੑ࣭ΛΔ ߦྻΛݻ༗ϕΫτϧͱݻ༗ͷू߹ʹղ ਖ਼ํߦྻɹͷݻ༗ϕΫτϧ ▸ ɹΛֻ͚ͨͱ͖εΧϥʔഒͷΈɼ͖มԽ͠ͳ͍ ▸ ͜ͷ࣌ͷεΧϥʔɹ͕ݻ༗
19 Av = λv A v v λ A = Vdiag(λ)V−1 ݻ༗ϕΫτϧΛ࿈݁ͯ͠ɹͱ͢Δͱ V ݻ༗ղɿ
ݻ༗ղ ݻ༗ղͷྲྀΕ ݻ༗ϕΫτϧɹɹɹΛฒͯͰ͖Δߦྻ ɹ 20 v1 , v2 Q [Av1
Av2] = [λv1 λv2] A [v1 v2] = [λv1 λv2] [ λ1 0 0 λ2 ] Q = [v1 v2] ͔Β A−1AQ = [ λ1 0 0 λ2 ] AQ = Q [ λ1 0 0 λ2 ]
ಛҟղ ݻ༗ղਖ਼ํߦྻͰͷΈఆٛ͞ΕΔ ɹݻ༗ϕΫτϧͷߦྻɹͱݻ༗ɹͷϕΫτϧ͔Β ಛҟղͰͭͷߦྻͰɹΛද͢ ɹඞͣ͠ਖ਼ํߦྻͰͳ͍ ɹର֯ߦྻͰର֯Λಛҟͱ͍͏ 21 A = Vdiag(λ)V−1
λ V ݻ༗ղɿ A A = UDV−T U = m × m A = m × n V = n × n D = m × n D D