@3:00 you talk about the psychological area about what makes good explanations. I am thinking to do my thesis around this topic Human Centered AI or user friendly AI. Can you direct me to some sources which helps me to understand these topics?
Hi, first of all thanks for the nice video! I have a question: Is the base value given by the shap explainer the appearance of a certain class (stroke in this case) in the data on which model was trained on?
Hi iam not familiar with this topic ,it is quite new for me ,but i like your way in presenting as usual,my question is this field normaly ml data science should know to anlysis the model .or this field could person work separately to just test and anlysis the module
Hi and thanks! It depends :) in my experience the person who build a model also works on interpreting it. This is to make sure the model works as intended - basically its a verification step after developing the algorithm. But I can imagine that also other people work on the XAI tasks especially in larger companies when larger models are used
By the way: more and more companies and frameworks also started to include interpretability tools. E. G. Tensorflow has the what-if tool or Google Has interpeetable AI for Google Cloud.
![The moment we stopped understanding AI [AlexNet]](/img/1.gif)
This is a great series. Eye opening indeed. Thanks. Look forward to new videos on other ML subjects.
such a great video! thank you for condensing it and also pointing out to directions where we can venture out.
Thanks for this video and playlist. It's really hard to get this info in such a simple manner.
Great intro and series to XAI - thanks
Thank you very much for making this playlist - absolute chamption!
Hi this series is extraordinarily good. Very well we explained. Thank you.
@3:00 you talk about the psychological area about what makes good explanations. I am thinking to do my thesis around this topic Human Centered AI or user friendly AI. Can you direct me to some sources which helps me to understand these topics?
Great video on a very important subject. Thanks!
This is a great tutorial! Many thanks!
Great explanation & Great series
Thank you for the amazing explaination
Hi, first of all thanks for the nice video! I have a question: Is the base value given by the shap explainer the appearance of a certain class (stroke in this case) in the data on which model was trained on?
Hi! Thanks :)
Yes exactely, the base value is calculated as average of the train dataset, as far as I know.
Good Stuff!
Hi iam not familiar with this topic ,it is quite new for me ,but i like your way in presenting as usual,my question is this field normaly ml data science should know to anlysis the model .or this field could person work separately to just test and anlysis the module
Hi and thanks! It depends :) in my experience the person who build a model also works on interpreting it. This is to make sure the model works as intended - basically its a verification step after developing the algorithm. But I can imagine that also other people work on the XAI tasks especially in larger companies when larger models are used
By the way: more and more companies and frameworks also started to include interpretability tools. E. G. Tensorflow has the what-if tool or Google Has interpeetable AI for Google Cloud.
cloud.google.com/explainable-ai here for example
Amazing 😍
Helpful
woowwwww!!
How can I get in touch with you to ask you some questions
deepfindr@gmail.com :)
@@DeepFindr Thanks, I' ve sent my problems to your email
Thx