Jason Grafft Enhancing Machine Learning and Data Visualization Pipelines with Isomorphisms λ C20 GE
Slides Machine learning and visualization operations typically require multiple transformations of input data. Entropy tends to increase with the number of transformations, expanding the likelihood of errant behavior that drains system resources and is challenging to identify because it is difficult to reproduce. Isomorpisms may be used to reduce the resources required and increase the explainability of transformations data undergo. We will discuss several example
|
|