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
[動画あり] 線形回帰を題材に汎用的な理解を身につける:座学編
Search
数理の弾丸
April 09, 2024
150
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
[動画あり] 線形回帰を題材に汎用的な理解を身につける:座学編
YouTube:
https://youtu.be/54pe6MDaGI0
数理の弾丸
April 09, 2024
More Decks by 数理の弾丸
See All by 数理の弾丸
grill-me(grilling):AI駆動開発を超助けてくれるSkill / grill me
mathbullet
0
35
Claude Code × autoresearch 実践
mathbullet
0
250
【動画あり】Transformer論文解説
mathbullet
0
490
RAG:チャットボットの能力を底上げする技術
mathbullet
0
300
ゼロから始める大規模言語モデル入門
mathbullet
0
300
[動画あり] AI入門特急コース
mathbullet
0
260
Featured
See All Featured
CoffeeScript is Beautiful & I Never Want to Write Plain JavaScript Again
sstephenson
162
16k
Sharpening the Axe: The Primacy of Toolmaking
bcantrill
46
3k
Navigating the moral maze — ethical principles for Al-driven product design
skipperchong
2
510
HDC tutorial
michielstock
2
830
Stewardship and Sustainability of Urban and Community Forests
pwiseman
0
500
Cheating the UX When There Is Nothing More to Optimize - PixelPioneers
stephaniewalter
287
14k
The Illustrated Guide to Node.js - THAT Conference 2024
reverentgeek
1
450
Context Engineering - Making Every Token Count
addyosmani
9
1.1k
The State of eCommerce SEO: How to Win in Today's Products SERPs - #SEOweek
aleyda
2
11k
Refactoring Trust on Your Teams (GOTO; Chicago 2020)
rmw
35
3.8k
Improving Core Web Vitals using Speculation Rules API
sergeychernyshev
21
1.6k
Rebuilding a faster, lazier Slack
samanthasiow
85
9.6k
Transcript
ຊεϥΠυΛ༻ͨ͠ղઆಈը ֓ཁཝͷϦϯΫ͔Β
εϐʔΧʔ ٢ా ژେใܥ%"*ίϯαϧݴޠֶɾࣗવݴޠॲཧ εϛε چఇେଔ3ˍ%ݚڀһԽֶܥ
ࠓճͷΰʔϧ ઢܗճؼϞσϧͷϝΧχζϜͱͦͷॴɾॴΛΔ ઢܗճؼϞσϧͷ1ZUIPO࣮ΠϝʔδΛ࣋ͭ ্هΛ௨ͯ͠ɺ৽ͨͳϞσϧΛʮ׆༻Ͱ͖Δঢ়ଶ·Ͱཧղ͢Δʯ ͱ͍͏͜ͱͷഽײ֮Λ௫Ή
৽ͨͳϞσϧΛʮ׆༻Ͱ͖Δঢ়ଶ·Ͱཧղ͢Δʯͱ͍͏͜ͱͷഽײ֮Λ௫Ήͷؾ࣋ͪ ઢܗϞσϧ ܾఆ χϡʔϥϧωοτ Lฏۉ๏ ϚϧίϑϞσϧ FUD ৽ͨͳٕज़Λशಘ͢Δඞཁੑৗʹ͋Δ
ࠓճͷΰʔϧ ઢܗճؼϞσϧͷϝΧχζϜͱͦͷॴɾॴΛΔ ઢܗճؼϞσϧͷ1ZUIPO࣮ΠϝʔδΛ࣋ͭ ্هΛ௨ͯ͠ɺ৽ͨͳϞσϧΛʮ׆༻Ͱ͖Δঢ়ଶ·Ͱཧղ͢Δʯ ͱ͍͏͜ͱͷഽײ֮Λ௫Ή
͜ͷಈըͰ৮Εͳ͍͜ͱ ֶशΞϧΰϦζϜͷৄࡉ ࠷খೋ๏ʹΑΔύϥϝʔλ࠷దԽ ઢܗճؼͷੜख๏ ਖ਼ଇԽɺϩδεςΟοΫճؼFUD
ճؼͱʁ
ճؼͱʁ ճؼ େখʹҙຯͷ͋ΔΛ༧ଌ͢Δઃఆ
ճؼͱʁ ճؼ େখʹҙຯͷ͋ΔΛ༧ଌ͢Δઃఆ Ωϟϯϖʔϯछผ ސ٬ಛੑ ༧ଌ ޮՌࢦඪ ΩϟϯϖʔϯͷޮՌ༧ଌ ྫ
ճؼͱʁ ճؼ େখʹҙຯͷ͋ΔΛ༧ଌ͢Δઃఆ Ωϟϯϖʔϯछผ ސ٬ಛੑ ༧ଌ ޮՌࢦඪ ΩϟϯϖʔϯޮՌͷࣄલγϛϡϨʔγϣϯ ൢଅޮՌͷߴ͍ސ٬ಛੑͷൃݟ
ΩϟϯϖʔϯͷޮՌ༧ଌ ྫ
ճؼͱʁ ճؼ େখʹҙຯͷ͋ΔΛ༧ଌ͢Δઃఆ Ωϟϯϖʔϯछผ ސ٬ಛੑ ༧ଌ ޮՌࢦඪ ΩϟϯϖʔϯޮՌͷࣄલγϛϡϨʔγϣϯ ൢଅޮՌͷߴ͍ސ٬ಛੑͷൃݟ
ΩϟϯϖʔϯͷޮՌ༧ଌ ྫ આ໌มʢ༧ଌͷใݯͱ͢Δʣ తมʢ༧ଌͷରͱ͢Δʣ
ઢܗճؼϞσϧͱʁ
ઢܗճؼϞσϧͱʁ ઢܗճؼ తมΛઆ໌มͷҰ࣍ࣜͰ༧ଌ͢ΔϞσϧ y = w1 x1 + w2
x2 + b
ઢܗճؼϞσϧͱʁ ઢܗճؼ తมΛઆ໌มͷҰ࣍ࣜͰ༧ଌ͢ΔϞσϧ y = w1 x1 + w2
x2 + b આ໌ม తม ༧ଌͷରͱ͢Δ ༧ଌͷใݯͱ͢Δ
ઢܗճؼϞσϧͱʁ ઢܗճؼ తมΛઆ໌มͷҰ࣍ࣜͰ༧ଌ͢ΔϞσϧ y = w1 x1 + w2
x2 + b આ໌ม తม ύϥϝʔλ ಛʹ ʮઆ໌มͷ֤ΛͲΕ͘Β͍༧ଌʹӨڹͤ͞Δ͔ʯΛද͢ w ༧ଌͷରͱ͢Δ ༧ଌͷใݯͱ͢Δ
ΠϝʔδʮઢͷͯΊʯ ސ٬ಛੑʢFH݄ͨΓߪങֹʣ ޮՌࢦඪ
ΠϝʔδʮઢͷͯΊʯ ސ٬ಛੑʢFH݄ͨΓߪങֹʣ ޮՌࢦඪ
ΠϝʔδʮઢͷͯΊʯ ސ٬ಛੑʢFH݄ͨΓߪങֹʣ ޮՌࢦඪ
ΠϝʔδʮઢͷͯΊʯ ސ٬ಛੑʢFH݄ͨΓߪങֹʣ ޮՌࢦඪ ઢͷܗΛܾΊΔͷ͖ͱย w ͕͖ΛܾΊɺ ͕ยΛܾΊΔ w ͯ·Γͷྑ͞
Ͱܾ·Δ w ֶश Λσʔλʹ߹Θͤͯௐ͢Δ͜ͱ w b w, b w, b
ࠓճͷΰʔϧ ઢܗճؼϞσϧͷϝΧχζϜͱͦͷॴɾॴΛΔ ઢܗճؼϞσϧͷ1ZUIPO࣮ΠϝʔδΛ࣋ͭ ্هΛ௨ͯ͠ɺ৽ͨͳϞσϧΛʮ׆༻Ͱ͖Δঢ়ଶ·Ͱཧղ͢Δʯ ͱ͍͏͜ͱͷഽײ֮Λ௫Ή
ઢܗճؼϞσϧͷॴͱॴ
y = w1 x1 + w2 x2 + b ઢܗճؼϞσϧͷॴͱॴ
ॴ Ϟσϧͷ͕ࣜγϯϓϧͰɺֶशޙͷঢ়ଶΛղऍ͍͢͠ ➡︎ ୯ʹ༧ଌثͱͯ͠͏͚ͩͰͳ͘ɺཁҼੳʹ׆༻͍͢͠ 2.5 0.8
y = w1 x1 + w2 x2 + b ઢܗճؼϞσϧͷॴͱॴ
ॴ Ϟσϧͷ͕ࣜγϯϓϧͰɺֶशޙͷঢ়ଶΛղऍ͍͢͠ ➡︎ ୯ʹ༧ଌثͱͯ͠͏͚ͩͰͳ͘ɺཁҼੳʹ׆༻͍͢͠ 2.5 0.8
ॴ ઢܗճؼϞσϧͷॴͱॴ આ໌มͱతมͷؒʹઢܗͷ͕ؔͳ͍߹ɺ ͏·͘ϞσϦϯά͢Δͷ͍͠
ઢܗճؼϞσϧͷॴͱॴ ॴ આ໌มͱతมͷؒʹઢܗͷ͕ؔͳ͍߹ɺ ͏·͘ϞσϦϯά͢Δͷ͍͠ ॴ Ϟσϧͷ͕ࣜγϯϓϧͰɺֶशޙͷঢ়ଶΛղऍ͍͢͠ ➡︎ ୯ʹ༧ଌثͱͯ͠͏͚ͩͰͳ͘ɺཁҼੳʹ׆༻͍͢͠
ࠓճͷΰʔϧ ઢܗճؼϞσϧͷϝΧχζϜͱͦͷॴɾॴΛΔ ઢܗճؼϞσϧͷ1ZUIPO࣮ΠϝʔδΛ࣋ͭ ্هΛ௨ͯ͠ɺ৽ͨͳϞσϧΛʮ׆༻Ͱ͖Δঢ়ଶ·Ͱཧղ͢Δʯ ͱ͍͏͜ͱͷഽײ֮Λ௫Ή