Custom Metric Examples
Common example use cases
Custom metrics are an extremely powerful tool to evaluate many dimensions of your ML model. From analyzing the business impact of your model to calculating a moving average, you can leverage custom metrics in many ways. Use this page as a guide on how you can create custom metrics tailored to your ML needs.
The following examples use the
arize-demo-fraud-use-casemodel to calculate business KPIs and other useful statistics using custom metrics.
We'll use this example to calculate the percentage of predictions with a fico score below 600. Assume that the lending model automatically rejects customers with scores below 600, so you want to track this metric to determine the health of inbound applications. We can do this by using a
FILTER (WHERE ... )clause, which allows us to apply the < 600 filter to only the numerator and not the denominator.
COUNT(*) FILTER (WHERE fico_score < 600) / COUNT(*) * 100
Weighted averages can be useful when different predictions are important to your business.
To do this, we'll calculate recall but weigh the predictions by the loan amount. This way, a false negative on more expensive loans will have a higher impact on the score than a false negative on a cheaper loan.
We can use a
WHEREclause to filter for the actual fraud cases, and a
FILTER (WHERE ... )clause to filter the numerator where the model incorrectly predicted
SUM(loan_amount) FILTER (WHERE categoricalPredictionLabel = 'fraud') /
WHERE categoricalActualLabel = 'fraud'
You can use
APPROX_COUNT_DISTINCTto get the cardinality of a dimension. In this example, we can calculate the average total loan amount per merchant.
SUM(loan_amount) / APPROX_COUNT_DISTINCT(merchant_ID)
Use natively supported performance metrics as a function that can take multiple arguments for enhanced performance monitoring flexibility. Additionally, create completely new metrics using conditionals and other logic shown below.
When a model has multiple predictions that are sent as features or tags, the
PRECISIONfunction allows you to specify the exact columns used. In this example, a user may have a model with multiple predictions, each one sent in as a tag or feature. The PRECISION() function allows you to specify different columns to use for predictions or actuals, using the
F1 score is a special case of F beta, where precision and recall are equally weighted. In some use cases, you may prefer to weigh precision more than recall, or vice versa, based on business needs. The
F_BETAfunction allows you to specify the
betaparameter to adjust this weighting. In this example, we want to weigh precision twice as much as recall, with a beta score of 0.5.
Custom metrics can be used to create performance metrics such as pinball loss. Pinball loss is useful for quantile forecasts. An example could be a weather prediction, such as an 80% chance of 20 degrees or colder weather.
Let y be the actuals and z be the prediction. α is the quantile, in this case, 0.8 (for 80% chance). The formula for pinball loss is as such.
L(y, z) = (y - z) * α if y >= z
= (z - y) * (1 - α) if z > y
This is what pinball loss would look like for a numeric model.
WHEN numericActualLabel >= numericPredictionLabel
THEN (numericActualLabel - numericPredictionLabel) * 0.8
ELSE (numericPredictionLabel - numericActualLabel) * (1 - 0.8)
Custom metrics are not just limited to predictions and actuals. For example, if multiple predictions at multiple quantiles are sent in as tags, such as a tag
prediction_temp_p80for the 80th percentile temperature prediction, pinball loss could look like this:
WHEN numericActualLabel >= prediction_temp_p80
THEN (numericActualLabel - prediction_temp_p80) * 0.8
ELSE (numericActualLabel - prediction_temp_p80) * (1 - 0.8)