Deep Learning in Scala 3 from Scratch - Alexey Novakov

Deep Learning is a class of Machine Learning algorithms which have become quite popular in the last years. At first glance, neural networks seem a bit daunting to approach yourself, however on closer look you discover that it’s just math with a lot of advanced engineering to make it perform well. While there are specialized programming languages for deep learning, they typically favor a style of language that is closer to mathematical notation instead of resembling regular programming languages. As we will discover in this talk, Scala as language with it’s ability to write highly concise and declarative code, is a perfect match to express neural network algorithms as well. We will leverage its features like type inference, operator overloading, extension methods, first class functions as well as Scala 3 optional braces syntax to implement Deep Learning Algorithms.