|
|
EVENT DETAILS |
In many real-world applications, data quality & curation & domain knowledge play a much larger role in building successful models than coming up with complex processing techniques & tweaking hyper-parameters.
Therefore, a data toolbox should enable users to understand both data & model, & not burden the practitioner with details of formatting plots, rote preprocessing steps & hyperparameters.
The dabl library is a first step in this direction. It provides automatic visualization routines & model inspection capabilities while automating away model selection.
This talk will introduce the dabl library & show how to use it to quickly create supervised models & identify modeling & data quality issues.
|
|
|
|
|
|