Hi
Assume that I create a model using AutoML with 50 features from the Vertex AI Feature Store and after training I found that from the 50 original features, 10 has a very low incidence over the model.
Looking to increase the accuracy, reduce the consumption of resources and increase the speed of the model:
Do I need to remove the 10 features from the Feature Store and deploy the model to the endpoint?
Should I retrain the model with the 40 features and deploy it to the end point?
Any comments more than appreciated
I wouldn't delete the features because there are some features that can be used to share, discover, and re-use ML features at scale, which can increase the velocity of developing and deploying new ML applications.