Computer shows why: Visualizing deep learning for fun and profit
Deep neural networks are the modern day Swiss-knife for image processing. I will review visualization methods for evaluating, debugging and interpreting the performance and decisions of deep networks. I will present some real-world usages that I applied in the medical domain. We will play hands-on with heatmap techniques using online notebooks, which you can later apply to your own data and models
Dr. Eyal Gruss is a machine learning researcher, consultant and teacher, working mainly in image and language processing. Eyal has a diverse industry background, including medical, financial, cyber, sensors, ads, web, real-estate and creative. Eyal holds a PhD in physics, and is also a new-media artist creating and using generative algorithms.