Model Mayhem

Will has access to a simple dataset with only one numerical feature and a target value.

He wants to build a couple of models to predict the target value using his dataset. He will try two different functions:

  • Model A: y = mx + b
  • Model B: y = x² + mx + b

Will split his dataset in two: 80% for training both models and 20% for testing them.

Which model will get the best results on Will's testing set?