MSE

Returns a table that displays the mean squared error of the prediction and response columns in a machine learning model.

Syntax

MSE ( targets, predictions ) OVER()

Arguments

targets

The model response variable, of type FLOAT.

predictions

A FLOAT input column that contains predicted values for the response variable.

Examples

Execute the MSE function on input table faithful_testing. The response variables appear in the column obs, while the prediction variables appear in the column prediction.

=> SELECT MSE(obs, prediction) OVER()
   FROM (SELECT eruptions AS obs,
                PREDICT_LINEAR_REG (waiting USING PARAMETERS model_name='myLinearRegModel') AS prediction
         FROM faithful_testing) AS prediction_output;
        mse        |                   Comments
-------------------+-----------------------------------------------
 0.252925741352641 | Of 110 rows, 110 were used and 0 were ignored
(1 row)