Configure the update frequency and specify other settings that control how an
adaptive model operates.
- In the navigation pane of Prediction Studio, click Models.
- Open the adaptive model that you want to edit, and then click the
- In the Model update frequency section, in the Update
model after every field, enter the number of responses that trigger the
When you update a model, ADM retrains the model with the specified number of
responses. After the update, the model becomes available to the client nodes for scoring
and the Pega Platform components that use the model.Note: You should
choose model update frequency so that for the average model instance, it would be
updated every 2 to 4 hours. For example, if a model instance (such as a web banner for
an offer) has 166 impressions per minute, then 20,000 (2*60*166=20,000) may be a good
number for the adaptive model rule in that channel. The default setting is 5000. If the
model update frequency is set too low, then the demand on the ADM nodes will be high. In
addition to this, a model update will also be done at least every 12 hours, to make sure
all recorded responses are regularly processed.
- In the Recording historical data section, specify if you want to
extract historical customer responses from adaptive models in your application.
- In the Advanced Settings section, choose the update scope:
- To use all received responses for each update cycle, click Use all
- To assign more weight to recent responses when updating a model, click
Use subset of responses.
- In the Monitor performance for the last field, enter the number
of weighted responses used to calculate the model performance that is used in monitoring.
The default setting is 0, which means that all historical
data is to be used in performance monitoring.
Data analysis binning
section, in the
field, enter a value between 0 and 1 that
determines the granularity of the predictor binning.
The higher the value, the more bins are created. The value represents a statistical
threshold that indicates when predictor bins with similar behavior are merged. The default
setting is 0.25. Note: This setting operates in conjunction with Grouping
minimum cases to control how predictor grouping is established. The fact
that a predictor has more groups typically increases the performance, however the model
might become less robust.
- In the Grouping minimum cases field, enter a value between 0 and
1 that determines the minimum percentage of cases per interval.
Higher values result in decreasing the number of groups, which can be used to
increase the robustness of the model. Lower values result in increasing the number of
groups, which can be used to increase the performance of the model. The default setting is
- In the Predictor selection section, in the Activate
predictors with a performance above field, enter a value between 0 and 1
that determines the threshold for excluding poorly performing predictors.
The value is measured as the Area Under the Curve (AUC) of the predictor as compared
to the outcome. A higher value results in fewer predictors in the final model. The minimum
performance of AUC is 0.5, therefore the value of the performance threshold should always
be set to at least 0.5. The default setting is 0.52.
- In the Group predictors with a correlation above field, enter a
value between 0 and 1 that determines the threshold for excluding correlated predictors.
The default setting is 0.8. Predictors that have a mutual correlation above this
threshold are considered similar, and only the best of those predictors are used for
adaptive learning. The measure is the correlation between the probabilities of positive
behavior of pairs of predictors.
- In the Audit history section, to capture adaptive model details
in the work object's history, select the Attach audit notes to work
object check box.
Caution: Enabling this setting causes significant performance overhead.