Representing Small Floats for Machine Learning
Jeffrey Sarnoff
The IEEE has a working group drafting a standard for Floating-point Arithmetic Formats in Machine Learning. This talk presents an overview of their collective implementation and covers design advantages found with using Julia.