CylancePERSONA Desktop training mode Skip Navigation

CylancePERSONA Desktop
training mode

After the
CylancePERSONA Desktop
agent is installed, it creates a behavioral model of the user's activity and uses the model to recognize deviations from the user's expected behavior. The model uses a score from 0 to 100 to rate the user's behavior, with 100 being a perfect match to the user's behavioral model.
CylancePERSONA Desktop
data models include:
  • Keystroke model: the way the user types on the keyboard
  • Mouse data: the way the user moves and clicks the mouse or trackpad
  • Process start data: the applications the user launches and when
  • Logon data: when the user logs on or when the user fails at logging on
When
CylancePERSONA Desktop
is first installed on a device, there is a period of time when the data models must be trained by the user's behavior. While in training mode, the
CylancePERSONA Desktop
policy should be in a passive mode, which means mitigation actions are disabled in the policy. This allows the model to train without triggering any authentication events.
While in training mode, if the user fails to log in to the device, an alert will still display in the console.
The
CylancePERSONA Desktop
model training should take one to two weeks for most users. Training depends on the level of activity the user provides. The more activity, the sooner the model training completes.
While in training mode, users will be prompted with two factor authentication daily. This is to ensure devices are protected, even during the training period.