Talk

I have no clue what my model is doing...

Thursday, May 23

12:35 - 13:05
RoomLasagna
LanguageEnglish
Audience levelIntermediate
Elevator pitch

Working in Data Science is already not simple when we are in controlled conditions with a fixed training and test dataset, but how can we cope with the real world? What if we have no control over the training data? What if there is no defined test set? What if we do not have ground truth? What if…

Abstract

In this talk we will explore complex real world situations in which a model may be deployed. We will examine some common real world scenarios that make Data Science considerably more complex than the training/test split we often encounter in tutorials; we will analyse Python libraries and techniques that will allow us to estimate how the model is performing in the real world, when we may not have true labels, or when only a subset of errors are found, or when the problem is ill defined and we need to implement an approximate solution.

TagsPandas, Machine-Learning, Science
Participant

Pietro Mascolo

Data nerd, martial artist, photographer.

I am currently working as a data professional in healthcare and health insurance at Carelon Global Solutions.