search

tiagosousagarcia / turing-data-science-workshop

repo for the hands on activities for the Turing Data Science Workshop
0 stars 0 forks source link

Do we really need Deep Learning? What is the rationale for the request? #5

Open tiagosousagarcia opened 2 years ago

tiagosousagarcia commented 2 years ago

Probably not. We can use whatever tools answer the question best. The PI usually uses relatively simple analyses like simple regressions. We can go a step up without going all the way to deep learning. Explainability is also important.

tiagosousagarcia commented 2 years ago

because it's the cutting edge method, should give the best results, it may be the best way of identifying the most significant factors