Related Posts

• 

  The place of data wrangling in the data value cycle

  Learn more about data wrangling and how we position it within the data value cycle.

• 

  Why we fail with machine learning, and how to avoid it.

  With 53-90 of project failure, chances are each of us will be confronted to it. If making mistakes is part of the journey, let’s at least learn how to avoid some common ones.

• 

  Stop obsessing about a perfect model

  A perfect model is often an illusion because of the ratio effort / impact, Model Quality Decay and Prediction Window. To be pragmatic, we should focus on models that are quick to update and implement.

• 

  Let’s stop misleading our users with meaningless classification rates!

  So, you think AI classifies well? How we end up misleading AI users about the efficiency of classification models