Monday
Start with the Regression script that we prepared during today's class. Replace the Auto MPG dataset provided in the tensorflow exercise with the Auto Imports dataset provided in the UCI Machine Learning Repository. Specify a model with the following target and features.
- highway-mpg (continuous)
- num-of-cylinders (categorical)
- engine-size (continuous)
- horsepower (continuous)
- curb-weight (continuous)
Specify and train both a multi-class linear regression and a multi-class DNN regression. Which of the two models produces a better loss metric (see this link for an explanation of the loss function). Produce a plot that supports your answer. Return to the remainder of the variables from the dataset and add additional continuous and categorical features with the intent of improving your loss metric. Produce a plot that demonstrates the value of your model. What is the best model your team was able to produce?