Something I see too often in both immature data science practitioners and data science teams is overcomplication. The desire to use complex neural networks when there isn't enough data to support them, or wanting to include all kinds of variables that don't actually add predictive value.
Ironically, using a simple …
read more