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
Tree Methods
Search
Sunmi Yoon
November 04, 2019
Technology
0
120
Tree Methods
Decision Tree, Random Forest를 dataitgirls3 학생들에게 가르치기 위해 만든 수업자료입니다.
Sunmi Yoon
November 04, 2019
Tweet
Share
More Decks by Sunmi Yoon
See All by Sunmi Yoon
데이터 분석가 채용 공고 읽는 방법
ysunmi0427
1
340
Deep down in classification 0.5 magic number
ysunmi0427
0
94
Confusion matrix
ysunmi0427
0
150
심슨의 역설
ysunmi0427
0
2.3k
회사는 어떤 사람을 데이터 분석가로 채용하고 싶어하는 것일까?
ysunmi0427
0
2.4k
Other Decks in Technology
See All in Technology
Green Tea Garbage Collector の今
zchee
PRO
2
380
KAGのLT会 #8 - 東京リージョンでGAしたAmazon Q in QuickSightを使って、報告用の資料を作ってみた
0air
0
190
自作LLM Native GORM Pluginで実現する AI Agentバックテスト基盤構築
po3rin
2
230
about #74462 go/token#FileSet
tomtwinkle
1
270
研究開発部メンバーの働き⽅ / Sansan R&D Profile
sansan33
PRO
3
20k
AWSにおけるTrend Vision Oneの効果について
shimak
0
110
生成AIで「お客様の声」を ストーリーに変える 新潮流「Generative ETL」
ishikawa_satoru
1
280
PLaMo2シリーズのvLLM実装 / PFN LLM セミナー
pfn
PRO
2
920
LLMアプリケーション開発におけるセキュリティリスクと対策 / LLM Application Security
flatt_security
7
1.7k
Goを使ってTDDを体験しよう!
chiroruxx
1
260
OCI Network Firewall 概要
oracle4engineer
PRO
1
7.7k
組織観点からIAM Identity CenterとIAMの設計を考える
nrinetcom
PRO
1
150
Featured
See All Featured
Performance Is Good for Brains [We Love Speed 2024]
tammyeverts
12
1.1k
Practical Tips for Bootstrapping Information Extraction Pipelines
honnibal
PRO
23
1.5k
Designing for humans not robots
tammielis
254
25k
Producing Creativity
orderedlist
PRO
347
40k
Templates, Plugins, & Blocks: Oh My! Creating the theme that thinks of everything
marktimemedia
31
2.5k
Testing 201, or: Great Expectations
jmmastey
45
7.7k
Rails Girls Zürich Keynote
gr2m
95
14k
The Language of Interfaces
destraynor
162
25k
Optimizing for Happiness
mojombo
379
70k
Building Applications with DynamoDB
mza
96
6.6k
[RailsConf 2023 Opening Keynote] The Magic of Rails
eileencodes
30
9.7k
A better future with KSS
kneath
239
17k
Transcript
Tree methods dataitgirls3 Instructor Sunmi Yoon
Decision Tree
Sex <= 0.5 gini = 0.473 samples = 891 value
= [549, 342] class = Survived Fare <= 26.269 gini = 0.306 samples = 577 value = [468, 109] class = Survived True Fare <= 48.2 gini = 0.383 samples = 314 value = [81, 233] class = Dead False gini = 0.226 samples = 415 value = [361, 54] class = Survived gini = 0.448 samples = 162 value = [107, 55] class = Survived gini = 0.447 samples = 225 value = [76, 149] class = Dead gini = 0.106 samples = 89 value = [5, 84] class = Dead
Sex <= 0.5 gini = 0.473 samples = 891 value
= [549, 342] class = Survived Fare <= 26.269 gini = 0.306 samples = 577 value = [468, 109] class = Survived True Fare <= 48.2 gini = 0.383 samples = 314 value = [81, 233] class = Dead False gini = 0.226 samples = 415 value = [361, 54] class = Survived gini = 0.448 samples = 162 value = [107, 55] class = Survived gini = 0.447 samples = 225 value = [76, 149] class = Dead gini = 0.106 samples = 89 value = [5, 84] class = Dead Root Node (ࡸܻ) Intermediate Node (о) Terminal Node, Leaf ()
Sex <= 0.5 gini = 0.473 samples = 891 value
= [549, 342] class = Survived Fare <= 26.269 gini = 0.306 samples = 577 value = [468, 109] class = Survived True Fare <= 48.2 gini = 0.383 samples = 314 value = [81, 233] class = Dead False gini = 0.226 samples = 415 value = [361, 54] class = Survived gini = 0.448 samples = 162 value = [107, 55] class = Survived gini = 0.447 samples = 225 value = [76, 149] class = Dead gini = 0.106 samples = 89 value = [5, 84] class = Dead അ ਤী ؘఠо ݻ ѐ ਤ೧ ח Ӓ ؘఠٜ যڃ ۄ߰ਸ оҊ ח
Sex <= 0.5 gini = 0.473 samples = 891 value
= [549, 342] class = Survived Fare <= 26.269 gini = 0.306 samples = 577 value = [468, 109] class = Survived True Fare <= 48.2 gini = 0.383 samples = 314 value = [81, 233] class = Dead False gini = 0.226 samples = 415 value = [361, 54] class = Survived gini = 0.448 samples = 162 value = [107, 55] class = Survived gini = 0.447 samples = 225 value = [76, 149] class = Dead gini = 0.106 samples = 89 value = [5, 84] class = Dead যڃ ӝળਵ۽ оӝܳ ೮ח (gini ژח entropy)
Sex <= 0.5 gini = 0.473 samples = 891 value
= [549, 342] class = Survived Fare <= 26.269 gini = 0.306 samples = 577 value = [468, 109] class = Survived True Fare <= 48.2 gini = 0.383 samples = 314 value = [81, 233] class = Dead False gini = 0.226 samples = 415 value = [361, 54] class = Survived gini = 0.448 samples = 162 value = [107, 55] class = Survived gini = 0.447 samples = 225 value = [76, 149] class = Dead gini = 0.106 samples = 89 value = [5, 84] class = Dead Terminal Nodeী بೠ ؘఠٜਸ যڌѱ ࠙ܨೡ Ѫੋ
sklearn Code
Impurity
Impurity ࢎѾաޖח Impurity (ࠛࣽب, ࠛഛपࢿ) ծইח ߑߨਵ۽ णפ. ࣽبо ૐоೞח
Ѫਸ فҊ Information gainۄҊ ೞӝب פ. য়ט ࢎѾաޖ ࠛࣽب ஏ ߑߨ , Gini Indexܳ ҕࠗפ.
Sex <= 0.5 gini = 0.473 samples = 891 value
= [549, 342] class = Survived Fare <= 26.269 gini = 0.306 samples = 577 value = [468, 109] class = Survived True Fare <= 48.2 gini = 0.383 samples = 314 value = [81, 233] class = Dead False gini = 0.226 samples = 415 value = [361, 54] class = Survived gini = 0.448 samples = 162 value = [107, 55] class = Survived gini = 0.447 samples = 225 value = [76, 149] class = Dead gini = 0.106 samples = 89 value = [5, 84] class = Dead G = d ∑ i=1 Ri ( 1 − m ∑ k=1 p2 ik) Step 1. gini = 0.473 ਸ ҅೧ যࠁࣁਃ Step 2. gini = 0.226 ਸ ҅೧ যࠁࣁਃ
https://imgur.com/n3MVwHW
Random Forest
ৈ۞ ܻٜਸ ‘ܰѱ’ ݅ٚ. https://www.researchgate.net/figure/Architecture-of-the-random-forest-model_fig1_301638643
https://community.alteryx.com/t5/Alteryx-Designer-Knowledge-Base/Seeing-the-Forest-for-the-Trees-An-Introduction-to-Random-Forest/ta-p/158062 bagging = bootstrap aggregating
Bagging ߓӦ(bagging) bootstrap aggregating ড۽, ࠗझە(bootstrap)ਸ ా೧ ઑӘঀ ܲ ള۲
ؘఠী ೧ ള۲ػ ӝୡ ࠙ܨӝ(base learner)ٜਸ Ѿ(aggregating)दఃח ߑߨ. ࠗझەۆ, য ള۲ ؘఠীࢲ ࠂਸ ೲਊೞৈ ਗ ؘఠࣇҗ э ӝ ؘఠࣇਸ ݅٘ח җਸ ݈ೠ. ߓӦਸ ా೧ ےؒ ನۨझܳ ള۲दఃח җ җ э ࣁ ױ҅۽ ೯ػ. 1. ࠗझە ߑߨਸ ా೧ Nѐ ള۲ ؘఠࣇਸ ࢤࢿೠ. 2. Nѐ ӝୡ ࠙ܨӝ(ܻ)ٜਸ ള۲दఅ. 3. ӝୡ ࠙ܨӝ(ܻ)ٜਸ ೞա ࠙ܨӝ(ےؒ ನۨझ)۽ Ѿೠ(ಣӐ ژח җ߈ࣻై ߑध ਊ). Wikipedia ےؒನۨझ > ߓӦਸ ਊೠ ನۨझ ҳࢿ
sklearn Code