The most likely concern with a one-feature, machine-learning model is high error due to:
A model with only one feature is unlikely to capture the true complexity of the data's underlying relationships, leading to systematic underfitting - i.e., high bias.
Currently there are no comments in this discussion, be the first to comment!