일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | ||
6 | 7 | 8 | 9 | 10 | 11 | 12 |
13 | 14 | 15 | 16 | 17 | 18 | 19 |
20 | 21 | 22 | 23 | 24 | 25 | 26 |
27 | 28 | 29 | 30 |
- Tomek links
- Inatance segmentation
- GAN
- Digital Pathology
- XAI
- OCSVM
- 국비지원교육
- Generative modeling
- Random Undersampling
- 거리 기반 이상탐지
- Fixed Learning
- PM4Py
- Data Imbalance
- 밀도 기반 이상탐지
- Text generation
- Grad-CAM
- 딥러닝
- Gausian Density Estimation
- Petri net
- multi modal
- Process Mining
- 프로세스 마이닝
- Condensed neares neighbor rule
- Meta heuristic
- One-Sided Selection
- 병리 AI
- auto encoder
- Clustering 기반 이상탐지
- SQL 데이터 분석 첫걸음
- Sequence data
- Today
- Total
목록XAI (2)
Create Opportunities

가장 많이 사용되는 eXplainable AI인 SHAP에 대한 내용이다. SHAP To explain the prediction of a data point "x" by computing the contribution of each feature to the prediction. SHAP은 Cooperative(=Coalitional) Game Theory에 기반이 된다. 쉽게 말해서, 각 변수들이 Player가 되는 것이다. 이들 각각이 y에 어떠한 기여를 했는가! 를 보고자 하는 것이다. SHAP을 이해하기 위해서는 shapley value를 먼저 알아야 한다. 이득을 공평히 나누자 는 것이 핵심 아이디어이다. SHAP의 수식은 아래와 같다. SHAP은 LIME은 지키지 못했던 세 가지 원칙을 지킨..

설명 가능한 인공지능을 배운다. The results of the solution can be understood by humans. The process of generating human-understand explanations. Understand the cause of a decision. The process of giving explanations to humans. 왜 필요할까? 1. Help building trust : 2. Promote safety : 3. Allow for contestability 신뢰 / 안정성 보장 / Black box 모델은 Rule based도 아니고, 일련의 추론 과정을 설명하지 않기 때문 인공지능이 설명력을 갖춘다면, 뭘 어떻게 갖출건데? 1. Simu..