Details
Machine Learning
BusinessLog integrates a Machine Learning system that progressively learns from the recorded logs, allowing it to determine whether a user's access is “expected” or “unexpected”.
BusinessLog integrates a Machine Learning system that progressively learns from recorded logs, enabling it to determine whether a logged event should be considered expected or unexpected based on established behavioural patterns.
These classifications are visible in the "Working Log" and "Access Logs". The "ML" column shows the status assigned by the model, which may take the following values:
-
Green: the event matches the expected behavioural pattern.
-
Orange: the event deviates from the expected pattern but relates to a recognised entity.
-
Red: the event deviates from the expected pattern and relates to an unregistered or unknown entity.
-
Blue clock: the event is associated with a known user and machine but occurred outside the usual hours learned by the model.
-
Red clock: the event occurred outside the standard working hours (defined in the "Setting").
NOTE:
During the initial learning phase, it is normal for many events to appear as orange or red, as the system has not yet fully learned from archived data. Classification accuracy improves over time.
These classifications are visible in the "Working Log" and "Access Logs". The "ML" column shows the status assigned by the model, which may take the following values: