Overview
That meetup was held by Data Science Tbilisi. I was talking about Interpretable Machine Learning. Actually, it was pretty new for me since it became very popular on that period. I found one book which was written by Christoph Molnar. It was pretty interesting. I had a little experience with interpretability things in Machine Learning like: ELI5, GradCam, etc but I didn’t know all the available methods well. Since I was limited to 1 hour I wasn’t able to present all the things so I showed only machine learning and not Deep Learning approaches. The most popular and widely used one in classical Machine Learning is shap. There are some use cases in Deep Learning also. The one I liked most is GanPaint. I learned a lot from that book. It’s pretty easy to read and doesn’t take too much time. I recommend you to read it.