Summary
In this chapter, we've learned the most important tips and best practices that we can apply during the implementation of a ML use case with BigQuery ML.
We've analyzed the importance of data preparation; we started looking at the data quality aspects; then, we've learned how we can easily split the data to get balanced training, validation, and test sets.
We then looked at how we can further improve a ML model's performance using BigQuery ML functions for feature engineering.
After that, we focused our attention on tuning hyperparameters. When we train a model, BigQuery ML allows us to choose different parameters, and these variables influence the training stage.
Finally, we have understood why it is so important to deploy BigQuery ML models on other platforms so that we get online predictions and satisfy near-real-time business scenarios.
Congratulations on finishing reading the book! You should now be able to use BigQuery ML for your business...
