Use the Auto MPG dataset auto-mpg.data, available at https://archive.ics.uci.edu/ml/datasets/Auto+MPG a) [Marks: 5%] Drop the car name...

 

Transcribed Image Text:

Use the Auto MPG dataset auto-mpg.data, available at https://archive.ics.uci.edu/ml/datasets/Auto+MPG a) [Marks: 5%] Drop the car name attribute. b) [Marks: 5%] Split the data into 80% training and 20% testing sets. c) [Marks: 10%] Attribute horsepower has 6 missing values. Fill these up with the average horsepower for the respective number of cylinders. d) [Marks: 10% ] Train a poor model to predict mpg in such a way that the model intentionally overfits the training data. e) [Marks: 10% ] Train a better model that does not overfit the training data. f) [Marks: 10% ] Generate a graph showing both models for d) and e); the graph should have in the x-axis the feature that has the highest correlation with the mpg label, and in the y-axis the mpg. g) [Marks: 10%] Then test both models d) and e) using the testing set of b), and report the MSE for both models d) and e). Submit: [1] your code in a file named auto.py, [2] the graph in a file called auto.jpeg, and [3] the MSEs for models d) and e) as a comment at the very top of your file auto.py. Use the Auto MPG dataset auto-mpg.data, available at https://archive.ics.uci.edu/ml/datasets/Auto+MPG a) [Marks: 5%] Drop the car name attribute. b) [Marks: 5%] Split the data into 80% training and 20% testing sets. c) [Marks: 10%] Attribute horsepower has 6 missing values. Fill these up with the average horsepower for the respective number of cylinders. d) [Marks: 10% ] Train a poor model to predict mpg in such a way that the model intentionally overfits the training data. e) [Marks: 10% ] Train a better model that does not overfit the training data. f) [Marks: 10% ] Generate a graph showing both models for d) and e); the graph should have in the x-axis the feature that has the highest correlation with the mpg label, and in the y-axis the mpg. g) [Marks: 10%] Then test both models d) and e) using the testing set of b), and report the MSE for both models d) and e). Submit: [1] your code in a file named auto.py, [2] the graph in a file called auto.jpeg, and [3] the MSEs for models d) and e) as a comment at the very top of your file auto.py.