Upgrade to Pro — share decks privately, control downloads, hide ads and more …

“Why Should I Trust You?” Explaining the Predic...

Sponsored · Your Podcast. Everywhere. Effortlessly. Share. Educate. Inspire. Entertain. You do you. We'll handle the rest. →
Avatar for Wei Lu Wei Lu
November 06, 2017
Science
0
58

“Why Should I Trust You?” Explaining the Predictions of Any Classifier

Papers We Love, Nov 2017

Avatar for Wei Lu

Wei Lu

November 06, 2017
Tweet

More Decks by Wei Lu

See All by Wei Lu
A Brief Introduction to the Basics of Game Theory
Avatar for Wei Lu weilu
0
140
Bundler vs npm
Avatar for Wei Lu weilu
0
540
Cryptocurrency
Avatar for Wei Lu weilu
1
240
Peatio - An open-source crypto currency exchange
Avatar for Wei Lu weilu
0
680
Bitcoin Applications in JavaScript
Avatar for Wei Lu weilu
0
89
Writing expressive tests with RSpec
Avatar for Wei Lu weilu
0
140

Other Decks in Science

See All in Science
Accelerated Computing for Climate forecast
Avatar for Jeongkyu Shin inureyes
PRO
0
160
凸最適化からDC最適化まで
Avatar for ざきまつ santana_hammer
1
370
俺たちは本当に分かり合えるのか？ ～ PdMとスクラムチームの “ずれ” を科学する
Avatar for Takeo Imai bonotake
2
2.1k
学術講演会中央大学学員会府中支部
Avatar for Y-h. Taguchi tagtag
PRO
0
360
データマイニング - ノードの中心性
Avatar for Y. Yamamoto trycycle
PRO
0
350
中央大学AI・データサイエンスセンター　2025年第6回イブニングセミナー　『知能とはなにか　ヒトとAIのあいだ』
Avatar for Y-h. Taguchi tagtag
PRO
0
140
知能とはなにかーヒトとAIのあいだー
Avatar for Y-h. Taguchi tagtag
PRO
0
190
コミュニティサイエンスの実践@日本認知科学会2025
Avatar for hayataka hayataka88
0
140
なぜ21は素因数分解されないのか？ - Shorのアルゴリズムの現在と壁
Avatar for Daiki Murata daimurat
0
340
データマイニング - グラフ埋め込み入門
Avatar for Y. Yamamoto trycycle
PRO
1
190
先端因果推論特別研究チームの研究構想と 人間とAIが協働する自律因果探索の展望
Avatar for Shohei SHIMIZU sshimizu2006
3
810
【論文紹介】Is CLIP ideal? No. Can we fix it?Yes! 第65回 コンピュータビジョン勉強会＠関東
Avatar for Shun Makino shun6211
5
2.4k

Featured

See All Featured
Money Talks: Using Revenue to Get Sh*t Done
Avatar for Nikki Halliwell nikkihalliwell
0
190
Optimising Largest Contentful Paint
Avatar for Harry Roberts csswizardry
37
3.6k
Pawsitive SEO: Lessons from My Dog (and Many Mistakes) on Thriving as a Consultant in the Age of AI
Avatar for David Carrasco davidcarrasco
0
92
A designer walks into a library…
Avatar for Paul Jervis Heath pauljervisheath
210
24k
HU Berlin: Industrial-Strength Natural Language Processing with spaCy and Prodigy
Avatar for Ines Montani inesmontani
PRO
0
280
GitHub's CSS Performance
Avatar for Jon Rohan jonrohan
1032
470k
I Don’t Have Time: Getting Over the Fear to Launch Your Podcast
Avatar for Joe Casabona jcasabona
34
2.7k
Test your architecture with Archunit
Avatar for Yoan thirion
1
2.2k
Digital Projects Gone Horribly Wrong (And the UX Pros Who Still Save the Day) - Dean Schuster
Avatar for UX Y'all uxyall
0
800
DBのスキルで生き残る技術 - AI時代におけるテーブル設計の勘所
Avatar for soudai sone soudai
PRO
64
52k
The AI Revolution Will Not Be Monopolized: How open-source beats economies of scale, even for LLMs
Avatar for Ines Montani inesmontani
PRO
3
3.2k
AI Search:
Where Are We & What Can We Do About It?
Avatar for Aleyda Solis aleyda
0
7.2k

Transcript

  1. “Why Should I Trust You?” Explaining the Predictions of Any

    Classifier Papers We Love Nov 2017
  2. None

  3. Can I trust the prediction?

  4. Can I trust the prediction?

  5. 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.

  6. Trust Prediction Model

  7. 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”

  8. Local fidelity

  9. Pick predictions: Submodular pick (SP)