Yes, it is possible to use the create/replace command of Bigquery ML to perform multiple regressions with a single model. This can be done by specifying multiple inputs and targets in the same CREATE MODEL statement. For example:
CREATE or REPLACE MODEL mymodel
OPTIONS (
model_type='linear_reg',
input_label_cols=['input1', 'input2', 'input3'],
output_label_cols=['target1', 'target2', 'target3']
) AS
SELECT input1, input2, input3, target1, target2, target3
FROM mytable
In this example, the model will be trained using three inputs (input1, input2, input3) and three targets (target1, target2, target3) from the same table.
Please start posting anonymously - your entry will be published after you log in or create a new account. This space is reserved only for answers. If you would like to engage in a discussion, please instead post a comment under the question or an answer that you would like to discuss
Asked: 2023-07-18 19:32:40 +0000
Seen: 11 times
Last updated: Jul 18 '23
Are IS_NUMERIC and IS_SPACES the same in BigQuery?
What is the method for retrieving a variable within the WHERE clause of BigQuery?
Is the GROUP BY executed following the operations executed in the SELECT statement in BigQuery?
How to retrieve lost messages in ZeroMQ publish/subscribe model?
How can weights be duplicated and altered using pytorch?
What are the steps to visualize a PCL Model accurately using ROS RVIZ?
What could be the reason for the low AUC and ROC scores of my significant model?