Evaluating the matrix factorization model
In this section, we'll evaluate the performances of the matrix factorization model that we trained in the previous section.
The evaluation stage of a matrix factorization model can be performed using the ML.EVALUATE BigQuery ML function or through the BigQuery UI.
Let's execute the following query to extract all the evaluation parameters that characterize the recommendation model that we've just trained:
SELECT * FROM ML.EVALUATE(MODEL `09_recommendation_engine.recommender`, ( SELECT * FROM `09_recommendation_engine.product_visits`));
The result of this query is shown in the following screenshot:
Figure 9.9 – The record that's been extracted from the evaluation of the matrix factorization model
The same information can be accessed by selecting the ML model from the BigQuery navigation menu and then accessing...