The art of storytelling for a data scientist

Data Scientists like other scientists are also experimenting, but with Data. Most of the machine learning algorithms are sensitive to the way you feed the data and the way you analyse the outcomes, making it a perfect ground for experimentation.

Experimenting implicitly means, there will be some trial and errors, some that will fail and some that will have an outcome which is valuable and few with breakthroughs.

Storytelling is an art, where you describe the following:

  1. The value of some valuable outcomes for the business/end users/research etc.
  2. The experiments that fail, so that people don’t retry those
  3. How you managed to fine tune the steps of your experiment that generated some value
  4. If all the experiments have failed so far, what have you learnt from them? And what is your plan of action?

A lot interviewers are also looking at this ability of a data scientist, than just the f1-score of your model!

1 Like