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 erroneous ...
Search
Kenji Saito
PRO
December 06, 2023
Business
170
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
起こりうる誤った推論/平均・分散・標準偏差・自由度 / Possible erroneous inferences, and mean, variance, standard deviation, degree of freedom
早稲田大学大学院経営管理研究科「企業データ分析」2023 冬の第3-4回で使用したスライドです。
Kenji Saito
PRO
December 06, 2023
More Decks by Kenji Saito
See All by Kenji Saito
文化的ループ (推論, エージェント) と思考プロセスの検証可能性 / Cultural Loops (Inference, Agents) and the Verifiability of Thought Processes
ks91
PRO
0
24
アカデミーキャンプ2026 夏 「ヤツらを究めろ!ウチらとヤツらの自由研究∼ シーズン2」DAY 3 / Academy Camp 2026 Summer: Master Them! Our Independent Research Project with Them—Season 2 DAY3
ks91
PRO
0
56
アカデミーキャンプ2026 夏 「ヤツらを究めろ!ウチらとヤツらの自由研究∼ シーズン2」DAY 1 / Academy Camp 2026 Summer: Master Them! Our Independent Research Project with Them—Season 2 DAY1
ks91
PRO
0
59
アカデミーキャンプ2026 夏 「ヤツらを究めろ!ウチらとヤツらの自由研究∼ シーズン2」DAY 2 / Academy Camp 2026 Summer: Master Them! Our Independent Research Project with Them—Season 2 DAY2
ks91
PRO
0
89
責任2.0 ∼ AI が主体となる世界で ∼ / Responsibility 2.0: In a World Where AI Takes Responsibilities
ks91
PRO
0
77
発表と総括 / Presentations and Summary
ks91
PRO
0
230
インシデントレスポンス演習II / Incident Response Exercise II
ks91
PRO
0
110
脱金融のフューチャー・デザイン / Future Design Beyond Finance
ks91
PRO
0
200
戦略と安全保障 / Strategy and Security
ks91
PRO
0
120
Other Decks in Business
See All in Business
ビジネスプロセスから始めるデータモデリング ファクトとディメンションの前に考えること
tenajima
4
970
AIネイティブな組織を問い直す
smiyawaki0820
15
6.9k
2026_1H_TransparencyReport_EN.pdf
mercari_inc
0
320
ゲームにおけるメディアミックス展開との向き合い方 -ヘブンバーンズレッドのメディアミックス企画の立て方-
gree_tech
PRO
0
110
医院継承G紹介資料_エムスリー株式会社
m3
0
1.8k
Playwright × AI Agent でE2Eテストはどう変わるか AI駆動テストの可能性と実用検証の結果 _0721
findy_eventslides
2
350
会社紹介 2026
xgene
0
290
スクラムマスターの観察眼 〜AIによる3days爆速キャッチアップと次の一手〜/The Scrum Master's Insight: Lightning-Fast 3-Day Catch-Up with AI and the Next Move
ikuodanaka
4
870
株式会社ナビタイムジャパン 会社紹介資料2026
navitimejapan
PRO
1
730
「エンジニアのための自己管理入門」のインセプションデッキ/Inception Deck of Self-Management beginner's guide book
ikuodanaka
2
560
セーフィー株式会社(Safie Inc.) 会社紹介資料
safie_recruit
7
460k
こだわりを手放したことで気づけたスクラムマスターとしての振る舞い
kawanotron
1
610
Featured
See All Featured
Practical Tips for Bootstrapping Information Extraction Pipelines
honnibal
25
2k
How GitHub (no longer) Works
holman
316
150k
[RailsConf 2023 Opening Keynote] The Magic of Rails
eileencodes
31
10k
Avoiding the “Bad Training, Faster” Trap in the Age of AI
tmiket
0
210
Facilitating Awesome Meetings
lara
57
7.1k
Building the Perfect Custom Keyboard
takai
2
840
Prompt Engineering for Job Search
mfonobong
0
410
Performance Is Good for Brains [We Love Speed 2024]
tammyeverts
12
1.8k
The Organizational Zoo: Understanding Human Behavior Agility Through Metaphoric Constructive Conversations (based on the works of Arthur Shelley, Ph.D)
kimpetersen
PRO
0
420
Mind Mapping
helmedeiros
PRO
1
320
The Psychology of Web Performance [Beyond Tellerrand 2023]
tammyeverts
49
3.5k
HU Berlin: Industrial-Strength Natural Language Processing with spaCy and Prodigy
inesmontani
PRO
0
650
Transcript
generated by Stable Diffusion XL v1.0 2023 3-4 (WBS) 2023
3-4 — 2023-12-07 – p.1/37
https://speakerdeck.com/ks91/collections/corporate-data-analysis-2023-winter 2023 3-4 — 2023-12-07 – p.2/37
( ) 1 11 30 • 2 11 30 (B
A ) • 3 12 7 • 4 12 7 • 5 12 14 6 12 14 t 7 12 21 2 ( ) t 8 12 21 2 ( ) t 9 1 11 P 10 1 11 11 1 18 12 1 18 13 1 25 14 1 25 W-IOI 2023 3-4 — 2023-12-07 – p.3/37
( 20 ) 1 • 2 R • 3 •
4 • 5 6 ( ) 7 (1) 8 (2) 9 R ( ) (1) 10 R ( ) (2) 11 R ( ) (1) 12 R ( ) (2) 13 GPT-4 14 GPT-4 15 ( ) LaTeX Overleaf 8 (12/14 ) / (2 ) OK / 2023 3-4 — 2023-12-07 – p.4/37
(B A ) 1 ( ) 2 (Wilcoxon-Mann-Whitney ) 2023
3-4 — 2023-12-07 – p.5/37
3 1 ( ) 2 ( ) 1 2 4
σ2 σ s2 s df 2023 3-4 — 2023-12-07 – p.6/37
2023 3-4 — 2023-12-07 – p.7/37
1. (1) (2) 2023 12 3 ( ) 23:59 JST
( ) Waseda Moodle (Q & A ) 2023 3-4 — 2023-12-07 – p.8/37
. . . . . . 13(12) 11(10) (12/5( )
) ( ) R ( ) (1) 2023 3-4 — 2023-12-07 – p.9/37
(1) ( ) 2 1 ⇒ 1 2 (2) 95%
⇒ ( 6 ) 11 2 95% (← be ambitious!) 5% (← 2 ) 2023 3-4 — 2023-12-07 – p.10/37
0 2 4 6 8 10 0.0 0.1 0.2 0.3
0.4 0.5 0.6 ᳨ฟᅇᩘ ☜⋡ 2023 3-4 — 2023-12-07 – p.11/37
W GitHub Gist ⇒ RStudio Git GitHub https://github.com/ks91/cda-demo “cda-demo” RStudio
Git Git Dropbox “cda-demo” 2023 3-4 — 2023-12-07 – p.12/37
Git Git ( GPL) GitHub Git ( ) RStudio pull
Git (OS ) Linux : ( OK) macOS : ( OK) Windows : https://gitforwindows.org OK https://github.com/ks91/cda-demo ( ) 2023 3-4 — 2023-12-07 – p.13/37
S ⇒ D R ( ) × * ÷ /
xy x^y ( ) √ x sqrt(x) (function; ) ( 1 , 2 ,. . .) sqrt(9) + sqrt(16) <- ( ) ( ) x <- x + 1 x 1 ( ) 2023 3-4 — 2023-12-07 – p.14/37
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$group == " ", "rank"]) ⇒ df group rank ra ( == ) 2023 3-4 — 2023-12-07 – p.15/37
“ D.R” ‘sum(. . .)’ # sum(df[df$group == " ",
"rank"]) ... <- 0 i <- 1 # while (i df ) { # if ((df i group ) == " ") { <- + (df i rank ) } i <- i + 1 } 2023 3-4 — 2023-12-07 – p.16/37
S ( ) ⇒ ↑ ( ) : Amazon PPT
Discord 2023 3-4 — 2023-12-07 – p.17/37
S 2 (A B ) ⇒ ( ) ( )
2023 3-4 — 2023-12-07 – p.18/37
S 1 1 1/6 5/6 1 1/6 ⇒ n 1
x ( n = 10 ) 2023 3-4 — 2023-12-07 – p.19/37
0 2 4 6 8 10 0.0 0.1 0.2 0.3
᳨ฟᅇᩘ ☜⋡ 2023 3-4 — 2023-12-07 – p.20/37
S A 5 B A ( ) A 5 B
A ( ) ( ) ( ) 1% 5 1/3 5 =0.41% 1% ⇒ n = 5 x = 5 x (1 3 )5 2023 3-4 — 2023-12-07 – p.21/37
0 1 2 3 4 5 0.0 0.1 0.2 0.3
᳨ฟᅇᩘ ☜⋡ 5% 2023 3-4 — 2023-12-07 – p.22/37
S 120g (120g ) 120g ⇒ 1kg 950g S 120g
· 120g 120g 2023 3-4 — 2023-12-07 – p.23/37
S ⇒ (p.84) χ2 H0 H0 [ ] RStudio Help
“chi” “chisq.test” 2023 3-4 — 2023-12-07 – p.24/37
χ2 1 6 (H0 ) # ( H_0 ) >
chisq.test(c(28, 33, 25, 35, 27, 32), p=c(1/6, 1/6, 1/6, 1/6, 1/6, 1/6)) # ( H_0 ) > chisq.test(c(42, 33, 25, 35, 27, 18), p=c(1/6, 1/6, 1/6, 1/6, 1/6, 1/6)) ‘chisq.test(. . .)’ ( ) ( ) ( ) ← R χ2 2023 3-4 — 2023-12-07 – p.25/37
K ⇒ 2023 3-4 — 2023-12-07 – p.26/37
. . . 2023 3-4 — 2023-12-07 – p.27/37
3 1 ( ) 2 ( ) 1 2 2023
3-4 — 2023-12-07 – p.28/37
100% 2 1 ( ( ) ) 2 ( (
) ) 2 1 1 α (α ) ( ) 2 β (β ) α n p 1 − β (power) 2023 3-4 — 2023-12-07 – p.29/37
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 2023 3-4 — 2023-12-07 – p.30/37
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 2023 3-4 — 2023-12-07 – p.31/37
4 σ2 σ s2 s df 2023 3-4 — 2023-12-07
– p.32/37
(parameter; ) ( ) µ, σ2, σ (statistic) ( )
x, s2, s (degree of freedom) ( ) df = n− k 2023 3-4 — 2023-12-07 – p.33/37
E ( p.106) 8 “ E.R” R ‘var(. . .)’,
‘sd(. . .)’ x sqrt(sd(x)^2*(length(x) - 1)/length(x)) 2023 3-4 — 2023-12-07 – p.34/37
2023 3-4 — 2023-12-07 – p.35/37
2. 1 2 (1) 1 2 (2) 2023 12 10
( ) 23:59 JST ( ) Waseda Moodle (Q & A ) (1) Discord 2023 3-4 — 2023-12-07 – p.36/37
2023 3-4 — 2023-12-07 – p.37/37