6 个机器学习可解释性框架!
Python学习与数据挖掘
共 2470字,需浏览 5分钟
· 2022-08-29
↓推荐关注↓
作者 | Moez Ali
来源 | DeepHub IMBA
如何学习的? 学到了什么? 针对一个特定输入为什么会做出如此决策? 决策是否可靠?
SHAP
基于随机森林模型的心脏病患者预测分类
![](https://filescdn.proginn.com/5a3cf4ca40276e756ac82b0fa14b143c/6fb4829c33ab80b736f754e06e583687.webp)
# install with pip
pip install shap
# install with conda
conda install -c conda-forge shap
![](https://filescdn.proginn.com/6653d2db459b0ee6019eefd1f1e1a7e9/ba98f1c66216a4ee55d16c4b3ff7b9ec.webp)
![](https://filescdn.proginn.com/6f844654f5364fc979d90353fa8a9c1a/757163752c7ddcac4b799c5e73a2983d.webp)
![](https://filescdn.proginn.com/6008f62898b7929b2ed6b46834f87655/d4c48e98873fd59cdec7fa50652f9697.webp)
LIME
![](https://filescdn.proginn.com/6f02b825380871792a694fec2fe015fd/142004e39cc4f41f7b612e561509e810.webp)
pip install lime
![](https://filescdn.proginn.com/5d403b77f77ac25b46d12b7261e3994b/b92b8b8565ae0f957968e1e4031a1d26.webp)
Shapash
![](https://filescdn.proginn.com/d832b794f19475275a82612604f5f056/f6509b7279e6025c44193203c6df023d.webp)
![](https://filescdn.proginn.com/763ed791cf42bb31fe73bf6e1377b16a/08534b937b412cf53471740e7f8295ca.webp)
![](https://filescdn.proginn.com/19c3450af1fec45da3affbe1d2cd08f0/6da5c0031e278b0bf995a4421fb322b3.webp)
InterpretML
![](https://filescdn.proginn.com/b199df35b7ed098855de71e53d276565/fcf1c955e252ca7b1035521b32fa0071.webp)
![](https://filescdn.proginn.com/b7cc8e16830127bc60f81c5aef8b7d84/8eb73958ffd83c873e1cf6fa03274b76.webp)
ELI5
scikit-learn XGBoost、LightGBM CatBoost Keras
检查模型参数并说明模型是如何全局工作的; 检查模型的单个预测并说明什么模型会做出这样的决定。
![](https://filescdn.proginn.com/adc970958d23acb45784bf51727ac6e7/38c688dfebdfeff41c7eba33480631a8.webp)
![](https://filescdn.proginn.com/c0a50e61fd1623c11f0c558071b7c397/a3e4185cae6326ff59568ae08f394c5f.webp)
OmniXAI
![](https://filescdn.proginn.com/f7739abc6a2f3c1b95bc3e4463533567/8c439d2fa0dceaf89a0a5da2235a899e.webp)
![](https://filescdn.proginn.com/8bf6e6c150a7721fdb89725f9847f313/b2959223ffc29bbcda7e3d9d577c3213.webp)
长按或扫描下方二维码,后台回复:加群,即可申请入群。一定要备注:来源+研究方向+学校/公司,否则不拉入群中,见谅!
(长按三秒,进入后台)
推荐阅读
评论