Abstract: In one embodiment, a method for authenticating a user with an electronic device is disclosed. The method incudes receiving digital sensor data from a motion sensor over a signal acquisition time period; deleting a beginning portion of the digital sensor data prior to the signal acquisition time period; suppressing signal components in the data associated with voluntary movement of the user; signal processing the suppressed digital sensor data to extract signal features representing neuro muscular tone of the user; tabulating the extracted signal features over periods of time into a feature vector table; executing a predictive model with the feature vector table; generating a numerical degree of matching level based on the feature vector table and the user parameter set; and making a determination to either authorize the user or not based on the numerical degree of matching level. The predictive model is trained by a user parameter set.

Abstract: A method of authenticating a user device in providing access to a computer resource, the method includes: extracting a plurality of device fingerprint records from an access log, each of the device fingerprint records associated with an unauthorized access of a computer resource; from each of the plurality of device fingerprint records, extracting a digital signature, each of the digital signatures comprising a plurality of session characteristics; determining, by a processing device from the digital signatures, a root signature pattern, the root signature pattern comprising a combination of values of one or more of the plurality of session characteristics; and identify a subsequent access request for the computer resource as an unauthorized access based on a comparison of a device fingerprint associated with the subsequent access request and the root signature pattern.

Abstract: In one embodiment, a method for authenticating a user with an electronic device is disclosed. The method incudes receiving digital sensor data from a motion sensor over a signal acquisition time period; deleting a beginning portion of the digital sensor data prior to the signal acquisition time period; suppressing signal components in the data associated with voluntary movement of the user; signal processing the suppressed digital sensor data to extract signal features representing neuro muscular tone of the user; tabulating the extracted signal features over periods of time into a feature vector table; executing a predictive model with the feature vector table; generating a numerical degree of matching level based on the feature vector table and the user parameter set; and making a determination to either authorize the user or not based on the numerical degree of matching level. The predictive model is trained by a user parameter set.