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
“Why Should I Trust You?” Explaining the Predic...
Search
Wei Lu
November 06, 2017
Science
0
52
“Why Should I Trust You?” Explaining the Predictions of Any Classifier
Papers We Love, Nov 2017
Wei Lu
November 06, 2017
Tweet
Share
More Decks by Wei Lu
See All by Wei Lu
A Brief Introduction to the Basics of Game Theory
weilu
0
120
Bundler vs npm
weilu
0
530
Cryptocurrency
weilu
1
210
Peatio - An open-source crypto currency exchange
weilu
0
670
Bitcoin Applications in JavaScript
weilu
0
84
Writing expressive tests with RSpec
weilu
0
130
Other Decks in Science
See All in Science
動的トリートメント・レジームを推定するDynTxRegimeパッケージ
saltcooky12
0
160
アナログ計算機『計算尺』を愛でる Midosuji Tech #4/Analog Computing Device Slide Rule now and then
quiver
1
200
MCMCのR-hatは分散分析である
moricup
0
390
メール送信サーバの集約における透過型SMTP プロキシの定量評価 / Quantitative Evaluation of Transparent SMTP Proxy in Email Sending Server Aggregation
linyows
0
960
Ignite の1年間の軌跡
ktombow
0
130
マウス肝炎ウイルス感染の遺伝子発現へのテンソル分解の適用によるSARS-CoV-2感染関連重要ヒト遺伝子と有効な薬剤の同定
tagtag
0
120
白金鉱業Meetup Vol.16_数理最適化案件のはじめかた・すすめかた
brainpadpr
4
1.8k
Symfony Console Facelift
chalasr
2
460
眼科AIコンテスト2024_特別賞_6位Solution
pon0matsu
0
420
Lean4による汎化誤差評価の形式化
milano0017
1
260
データベース02: データベースの概念
trycycle
PRO
2
770
How To Buy, Verified Venmo Accounts in 2025 This year
usaallshop68
2
150
Featured
See All Featured
It's Worth the Effort
3n
185
28k
Thoughts on Productivity
jonyablonski
69
4.7k
Embracing the Ebb and Flow
colly
86
4.8k
The Cost Of JavaScript in 2023
addyosmani
51
8.6k
Keith and Marios Guide to Fast Websites
keithpitt
411
22k
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
229
22k
Rails Girls Zürich Keynote
gr2m
95
14k
Scaling GitHub
holman
460
140k
RailsConf 2023
tenderlove
30
1.1k
The World Runs on Bad Software
bkeepers
PRO
70
11k
Why You Should Never Use an ORM
jnunemaker
PRO
58
9.5k
Adopting Sorbet at Scale
ufuk
77
9.5k
Transcript
“Why Should I Trust You?” Explaining the Predictions of Any
Classifier Papers We Love Nov 2017
None
Can I trust the prediction?
Can I trust the prediction?
Trust • Trusting a prediction: whether a user trusts an
individual prediction sufficiently to take some action based on it. • Trusting a model: whether the user trusts a model to behave in reasonable ways if deployed.
Trust Prediction Model
LIME: Local Interpretable Model-agnostic Explanations Local: global model can be
complicated and hard to approximate Interpretable: because human Model-agnostic: works on any black box Explanation: back to “trust”
Local fidelity
Pick predictions: Submodular pick (SP)