Description
PyData Seattle 2017
Many companies implementing Machine Learning (ML) have learned that noise and other errors in the data set can cause stability issues resulting in time loss and headache.
Robust algorithms are under-appreciated, particularly by people new to data analysis. This talk will review the basic idea of robust or non-parametric algorithms and look at some of the more important named algorithms, as well as looking at how to apply the philosophy of robustness to any problem.