Simple models work. Linear models work. No need for deep learning or complex ensembles, you can often keep it simple. In this talk I'll discuss and demonstrate some winning tricks that you can apply on simple, even linear models.
It's not like deep learning isn't amazing, it is, but the hype is distracting people from great ideas. You may be doing yourself short if you pay too much attention to them. You can often win with simpler models that have properties that are much nicer in situations beyond a jupyter notebook (production).
I'll be discussing approaches that are dated, but useful. My goal is to lure you back to the domain of boring, old but ultimately simpler models that can win the day. I'll share some war-stories from my last 4 years of consulting work as well as some hard learned lessons.
The following topics will be discussed: