Abstract: Aspects and embodiments herein pertain to global ML prediction applied in a data security context. In some examples, a global prediction pertaining to a data item as a whole (the ‘original’ data item) is generated using an ML predictor. A data element is masked by removing it from the data item, or obscuring it in some way (e.g., through modification or replacement). The same ML predictor is applied to the perturbed data item. A change in global confidence between the original data item and the perturbed data item denotes local significance of the individual data item (or subset of data items) that have been masked. Other significance indicators (such as start and end markers of significant sections) are also considered. Other methods of estimating local significance are considered. In some examples, one or more such methods are used to generate training data for training a self-interpretation ML model.

Abstract: A user may be authenticated to access an account, computing device, or other resource based on the user's gaze pattern and neural or other physiological response(s) to one or more images or other stimuli. When the user attempts to access the resource, a computing device may obtain login gaze tracking data and measurement of a physiological condition of the user at the time that the user is viewing an image or other stimulus. Based on comparison of the login gaze tracking data and the measurement of the physiological condition to a model, the computing device can determine whether to authenticate the user to access the resource.

Abstract: A user may be authenticated to access an account, computing device, or other resource based on the user's gaze pattern and neural or other physiological response(s) to one or more images or other stimuli. When the user attempts to access the resource, a computing device may obtain login gaze tracking data and measurement of a physiological condition of the user at the time that the user is viewing an image or other stimulus. Based on comparison of the login gaze tracking data and the measurement of the physiological condition to a model, the computing device can determine whether to authenticate the user to access the resource.