Don't force imputation of numerical missing values in AutoML
yashpuranik
Partner, Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Neuron, Dataiku DSS Adv Designer, Registered, Neuron 2022, Neuron 2023 Posts: 69 Neuron
The visual ML Feature handling forces numerical columns to either drop rows with missing values or impute them.
There are algorithms (like Histogram based gradient boosting) that are able to natively handle missing values. Suggest having an option to keep missing values as is with compatible algorithms, especially to enable custom user implemented algorithms, even if none currently exist natively in the Visual ML selection.
Tagged:
Comments
-
yashpuranik Partner, Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Neuron, Dataiku DSS Adv Designer, Registered, Neuron 2022, Neuron 2023 Posts: 69 Neuron
