Distil’s machine learning feature adds yet another layer of protection by using behavioral modeling and pattern recognition to parse out bad users and bots from good traffic. Machine learning settings let you set both the threshold and protective action taken when Distil perceives a threat.
NOTE: This feature is available to domains that have been migrated to the latest version of Distil.
The machine learning graph displays a predictive estimate of the number of blocked requests based on your account’s machine learning threshold. It shows how users are blocked as you increase or decrease the machine learning scale.
NOTE: The graph shows data for the entire domain and is based on historical data covering the past seven days.
Use the Action dropdown list to set the responsive action taken when Distil’s machine learning suspects a malicious bot is attempting access.
The machine learning scale sets threshold aggressiveness.
NOTE: This scale only affects requests identified by machine learning. Distil’s primary web security protection identifies malicious bots no matter how aggressive your machine learning is set.
- Less Aggressive – Slide the control left to decrease the machine learning threshold.
Caution: Doing so potentially lets through bots with human-like behavior.
- More Aggressive – Slide the control right to increase the machine learning threshold.
Caution: Doing so potentially blocks human requests that show bot-like behavior.