Low-Code AI

Chapter 8. Improving Custom Model Performance

In Chapters 6 and 7, you learned how to prepare data and build custom models using SQL, BigQuery ML, and Python using scikit-learn and Keras. You will revisit those tools in this chapter with an eye toward additional feature engineering and hyperparameter tuning. In contrast to previous chapters, you will start with prepared data and an already trained model and work to improve from there. If you are confused when exploring the code for the previously built models or the user interface for BigQuery, please revisit the discussions in Chapters 6 and 7.

The Business Use Case: Used Car Auction Prices

Your goal in this project will be to improve the performance of an ML model trained to predict the auction price of used cars. The initial model is a linear regression model built in scikit-learn and does not quite meet your business goals. You will ultimately explore using tools in scikit-learn, Keras, and BigQuery ML to improve your model performance via feature engineering and hyperparameter tuning.

The dataset used for training the linear regression model has been supplied to you as CSV files. These datasets have been cleaned (missing and incorrect values have been remedied appropriately), and the code that was used to build the scikit-learn linear regression model has also been provided. Your teammate who trained the linear regression model has shared some notes with you on model performance and their initial explorations into using ...