Open the Black Box: an Introduction to Model Interpretability with LIME and SHAP - Kevin Lemagnen

PyData NYC 2018
What's the use of sophisticated machine learning models if you can't interpret them? This workshop covers two recent model interpretability techniques that are essentials in your data scientist toolbox: LIME and SHAP. You will learn how to apply these techniques in Python on a real-world data science problem.
===
www.pydata.org
PyData is an educational program of NumFOCUS, a 501(c)3 non-profit organization in the United States. PyData provides a forum for the international community of users and developers of data analysis tools to share ideas and learn from each other. The global PyData network promotes discussion of best practices, new approaches, and emerging technologies for data management, processing, analytics, and visualization. PyData communities approach data science using many languages, including (but not limited to) Python, Julia, and R.
PyData conferences aim to be accessible and community-driven, with novice to advanced level presentations. PyData tutorials and talks bring attendees the latest project features along with cutting-edge use cases. 00:00 Welcome!
00:10 Help us add time stamps or captions to this video! See the description for details.
Want to help add timestamps to our CZcams videos to help with discoverability? Find out more here: github.com/numfocus/CZcamsVi...

[@maheshmm2]

3:33 : github and colab links to code
5:20 : why is it important? Data bias
12:42: Explain like i'm 5
14:19 : Introduction to Interoperability (Jupyter code)
15:38 : sklearn.compose import column.transformer
20:55 : train, test
21.41: white box models, logistic regression
30:00: probability , score explained.
35.27 : Decision tree
36.28 : LIME
45:17 : LIME API
46:00: Random Forest
60:23 : SHAP
64:05 : SHAP API
75:31 : no tabular data
83:00 : Conclusion

[@narotian]

I appreciate that well explained.

[@bryanparis7779]

55:58 According to LIME, do these blue contributions of features really sum up to the probability of 0.71, if we show all contributions? Similarly, the orange ones are we sure that sum up to 0.29? I have examples of making me confused about this....