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
起こりうる誤った推論/平均・分散・標準偏差・自由度 / Possible false infe...
Search
Kenji Saito
PRO
December 06, 2024
Technology
0
120
起こりうる誤った推論/平均・分散・標準偏差・自由度 / Possible false inferences, means, variances, standard deviations and degrees of freedom
早稲田大学大学院経営管理研究科「企業データ分析」2024 冬の第3-4回で使用したスライドです。
Kenji Saito
PRO
December 06, 2024
Tweet
Share
More Decks by Kenji Saito
See All by Kenji Saito
ロボットを雰囲気(ヴァイブ)でプログラミングするこどもたち / Children Vibe-Programming Robots
ks91
PRO
0
7
アカデミーキャンプ 2025 SuuuuuuMMeR「燃えろ!!ロボコン」 / Academy Camp 2025 SuuuuuuMMeR "Burn the Spirit, Robocon!!" DAY 3
ks91
PRO
0
13
アカデミーキャンプ 2025 SuuuuuuMMeR「燃えろ!!ロボコン」 / Academy Camp 2025 SuuuuuuMMeR "Burn the Spirit, Robocon!!" DAY 2
ks91
PRO
0
22
アカデミーキャンプ 2025 SuuuuuuMMeR「燃えろ!!ロボコン」 / Academy Camp 2025 SuuuuuuMMeR "Burn the Spirit, Robocon!!" DAY 1
ks91
PRO
0
140
未来へのフォワードキャスト / Forward Cast to the Future
ks91
PRO
0
76
発表と総括 / Presentations and Summary
ks91
PRO
0
57
サイバーフィジカル社会、金融の未来とアイデアソン / Cyber Physical Society, Future of Finance, and Ideathon
ks91
PRO
0
75
マニフェスト: 人類の知のフロンティアに向けた拡張的足場へ / Manifesto: Toward Expansive Scaffolding for Humanity's Knowledge Frontier
ks91
PRO
0
36
続・スマートコントラクトと分散ファイナンス / Smart Contracts and Decentralized Finance, Continued
ks91
PRO
0
75
Other Decks in Technology
See All in Technology
生成AI導入の効果を最大化する データ活用戦略
ham0215
0
140
Vision Language Modelと自動運転AIの最前線_20250730
yuyamaguchi
4
1.3k
風が吹けばWHOISが使えなくなる~なぜWHOIS・RDAPはサーバー証明書のメール認証に使えなくなったのか~
orangemorishita
15
5.7k
マルチプロダクト×マルチテナントを支えるモジュラモノリスを中心としたアソビューのアーキテクチャ
disc99
1
450
AI関数が早くなったので試してみよう
kumakura
0
270
OPENLOGI Company Profile for engineer
hr01
1
38k
LLMをツールからプラットフォームへ〜Ai Workforceの戦略〜 #BetAIDay
layerx
PRO
1
950
生成AI時代におけるAI・機械学習技術を用いたプロダクト開発の深化と進化 #BetAIDay
layerx
PRO
1
1.1k
九州の人に知ってもらいたいGISスポット / gis spot in kyushu 2025
sakaik
0
140
僕たちが「開発しやすさ」を求め 模索し続けたアーキテクチャ #アーキテクチャ勉強会_findy
bengo4com
0
2.3k
[OCI Technical Deep Dive] OracleのAI戦略(2025年8月5日開催)
oracle4engineer
PRO
1
120
✨敗北解法コレクション✨〜Expertだった頃に足りなかった知識と技術〜
nanachi
1
680
Featured
See All Featured
[RailsConf 2023] Rails as a piece of cake
palkan
56
5.7k
Navigating Team Friction
lara
188
15k
Building Better People: How to give real-time feedback that sticks.
wjessup
367
19k
Fashionably flexible responsive web design (full day workshop)
malarkey
407
66k
Adopting Sorbet at Scale
ufuk
77
9.5k
Visualizing Your Data: Incorporating Mongo into Loggly Infrastructure
mongodb
47
9.6k
Java REST API Framework Comparison - PWX 2021
mraible
33
8.8k
Unsuck your backbone
ammeep
671
58k
Reflections from 52 weeks, 52 projects
jeffersonlam
351
21k
Being A Developer After 40
akosma
90
590k
Responsive Adventures: Dirty Tricks From The Dark Corners of Front-End
smashingmag
251
21k
Building an army of robots
kneath
306
45k
Transcript
Corporate data analysis — generated by Stable Diffusion XL v1.0
2024 3-4 (WBS) 2024 3-4 — 2024-12-09 – p.1/32
https://speakerdeck.com/ks91/collections/corporate-data-analysis-2024-winter 2024 3-4 — 2024-12-09 – p.2/32
( ) 1 12 2 • 2 12 2 (B
A ) • 3 12 9 • 4 12 9 • 5 12 16 6 12 16 t 7 12 23 2 ( ) t 8 12 23 2 ( ) t 9 1 6 P 10 1 6 11 1 20 12 1 20 13 1 27 14 1 27 W-IOI 2024 3-4 — 2024-12-09 – p.3/32
( 20 25 ) 1 (20 ) • 2 R
( 55 ) • 3 (32 ) • 4 (14 ) • 5 ( Git) (22 ) • 6 ( ) (24 ) • 7 (1) (25 ) • 8 (2) (25 ) • 9 R ( ) (1) — Welch (17 ) • 10 R ( ) (2) — (21 ) • 11 R ( ) (1) — (15 ) • 12 R ( ) (2) — (19 ) • 13 GPT-4 (19 ) • 14 GPT-4 (29 ) • 15 ( ) LaTeX Overleaf (40 ) • 8 (12/16 ) / (2 ) OK / 2024 3-4 — 2024-12-09 – p.4/32
(B A ) 1 ( ) 2 (Wilcoxon-Mann-Whitney ) 2024
3-4 — 2024-12-09 – p.5/32
3 1 ( ) 2 ( ) 1 2 4
σ2 σ s2 s df 2024 3-4 — 2024-12-09 – p.6/32
2024 3-4 — 2024-12-09 – p.7/32
1. (1) (2) 2024 12 5 ( ) 23:59 JST
( ) Waseda Moodle (Q & A ) 2024 3-4 — 2024-12-09 – p.8/32
. . . . . . 17 17 (12/6( )
) ( ) . . . 5 ( . . .) ( ) . . . 5 ( ) . . . 6 (2 ) R ^^; ( ) ( ) ( ) (1) 2024 3-4 — 2024-12-09 – p.9/32
(1) sqrt ( ) ^2 ( ) ( 3 )
(2) ( ) (3) 1 2 p (p WMW ) 16 ( ) 6 50% (← be ambitious!) 5% (← 2 ) 2024 3-4 — 2024-12-09 – p.10/32
( 50% ) 0 5 10 15 0.00 0.05 0.10
0.15 0.20 0.25 ᳨ฟᅇᩘ ☜⋡ 6 . . . 2024 3-4 — 2024-12-09 – p.11/32
H ⇒ ( : ) ( : ) 2024 3-4
— 2024-12-09 – p.12/32
I ( ) ex (P ) ( 5% or 1%
) ⇒ ⇒ ⇒ ⇒ ( ) 2024 3-4 — 2024-12-09 – p.13/32
Git Git ( GPL) GitHub Git ( ) RStudio pull
( ) Git (OS ) Linux : ( OK) macOS : Xcode (Apple ; App Store ) Windows : https://gitforwindows.org OK https://github.com/ks91/cda-demo ( ) 2024 3-4 — 2024-12-09 – p.14/32
U R ⇒ ( ) ( ) ( ) (
) : https://qiita.com/morayl/items/7d3a06d79fe2ab542b39 2024 3-4 — 2024-12-09 – p.15/32
H R WBS ⇒ R R 2024 3-4 — 2024-12-09
– p.16/32
R ⇒ D R ( ) × * ÷ /
xy x^y ( ) √ x sqrt(x) (function; ) ( 1 , 2 ,. . .) sqrt(9) + sqrt(16) ( ) <- ( ) ( ) x <- x + 1 x 1 ( ) 2024 3-4 — 2024-12-09 – p.17/32
R ( ) ⇒ D R " " T (true;
) F (false; ) c( 1 , 2 , . . .) ( ) 10 x x[1:3] 1 3 ( 1:100 1 100 ) a.b . . . . R R Source “ D.R” . . . ra <- sum(df[df$group == " ", "rank"]) ⇒ df group rank ra ( == ) 2024 3-4 — 2024-12-09 – p.18/32
“ D.R” ‘sum(. . .)’ # sum(df[df$group == " ",
"rank"]) ... <- 0 i <- 1 # while (i df ) { # if ((df i group ) == " ") { <- + (df i rank ) } i <- i + 1 } 2024 3-4 — 2024-12-09 – p.19/32
O R web ⇒ ( ) https://okumuralab.org/∼okumura/stat/ L A TEX(
/ ) ( ) R R (RStudio) ( ) 2024 3-4 — 2024-12-09 – p.20/32
N AI GPT ⇒ GPT Python R R ( (
) ) 2024 3-4 — 2024-12-09 – p.21/32
H ⇒ 2024 3-4 — 2024-12-09 – p.22/32
3 1 ( ) 2 ( ) 1 2 2024
3-4 — 2024-12-09 – p.23/32
100% 2 1 ( ( ) ) 2 ( (
) ) 2 1 1 α (α ) ( ) 2 β (β ) α n p 1 − β (power) 2024 3-4 — 2024-12-09 – p.24/32
3 “ .R” 3 bidist(n, p) : binull(n, p) :
5% bidistg(n, p0, p) : p0 p ( n = 20) p = 0.6 p = 0.2 p = 0.4 p = 0.8 p = 0.7 2024 3-4 — 2024-12-09 – p.25/32
0 5 10 15 20 0.00 0.05 0.10 0.15 0.20
᳨ฟᅇᩘ ☜⋡ 0 5 10 15 20 0.00 0.05 0.10 0.15 0.20 p = 0.2 2024 3-4 — 2024-12-09 – p.26/32
4 σ2 σ s2 s df 2024 3-4 — 2024-12-09
– p.27/32
(parameter; ) ( ) µ, σ2, σ (statistic) ( )
x, s2, s (degree of freedom) ( ) df = n− k 2024 3-4 — 2024-12-09 – p.28/32
E ( p.106) 8 “ E.R” R ‘var(. . .)’,
‘sd(. . .)’ x sqrt(sd(x)^2*(length(x) - 1)/length(x)) 2024 3-4 — 2024-12-09 – p.29/32
2024 3-4 — 2024-12-09 – p.30/32
2. 1 2 (1) 1 2 (2) 2023 12 12
( ) 23:59 JST ( ) Waseda Moodle (Q & A ) (1) Discord 2024 3-4 — 2024-12-09 – p.31/32
2024 3-4 — 2024-12-09 – p.32/32