DAT Solutions

Presentation and open discussions. “Introducing TensorFlow Probability” by David Molina, Data Scientist

Abstract: David will be presenting probabilistic modeling using TensorFlow. If we are classifying several thousands of images it's fine if we have some miss-classification detecting a cat or a dog but if we have our model making important decisions, for example, a miss-classification in a self-driving car or giving a medical diagnostic. Probabilistic programming allows us to deal with real-world uncertainty. Let's explore with TensorFlow probability how we can improve predictions adding randomness into our models.

8.15 pm – 9.00 pm: Coding and learning session. Bring your Python development laptop!