Abstract: A system for conducting a security recognition task, the system comprising a memory configured to store a model and training data including auxiliary information that will not be available as input to the model when the model is used as a security recognition task model for the security recognition task. The system further comprising one or more processors communicably linked to the memory and comprising a training unit and a prediction unit. The training unit is configured to receive the training data and the model from the memory and subsequently provide the training data to the model, and train the model, as the security recognition task model, using the training data to predict the auxiliary information as well as to perform the security recognition task, thereby improving performance of the security recognition task. The prediction unit is configured to use the security recognition task model output to perform the security recognition task while ignoring the auxiliary attributes in the model output.

Abstract: A system of one embodiment provides for efficient grid-estimation of spherical geo-probability function. The system includes a memory and a processor. The system accesses data, wherein the data includes training points and each training point includes a latitude value and a longitude value. The system also generates one or more grid points around each training point in the data. The system calculates a probability value for each grid point in the plurality of grid points using a probability density function. The system also combines each grid point into a geo-grid. The systems stores the geo-grid. In some embodiments, the system combines each grid point into a geo-grid by adding a probability value of a first grid point to a probability value of second grid point.

Abstract: This disclosure describes techniques and mechanisms for authenticating user device(s) by ensuring that two user devices accessing the same online service are co-located, while protecting private information associated with a user's network landscape and/or Bluetooth device(s). The techniques may ensure that a second factor authentication device is in the same location as a first factor access device, and that a first factor access device is in the same location as it was during previous access attempts).

Abstract: A system for conducting a security recognition task, the system comprising a memory configured to store a model and training data including auxiliary information that will not be available as input to the model when the model is used as a security recognition task model for the security recognition task. The system further comprising one or more processors communicably linked to the memory and comprising a training unit and a prediction unit. The training unit is configured to receive the training data and the model from the memory and subsequently provide the training data to the model, and train the model, as the security recognition task model, using the training data to predict the auxiliary information as well as perform the security recognition task, thereby improving performance of the security recognition task. The prediction unit is configured to use the security recognition task model output to perform the security recognition task while ignoring the auxiliary attributes in the model output.

Abstract: Statistical properties of known malware distributions may be used to improve estimates of malware detection metrics such as a base rate of malicious events in a target environment or missed detections (also referred to as false negatives). In particular, numerous synthetic sample distributions may be generated based on the statistical properties of a base data set and/or additional observed data, and used to identify malware distributions that produce overall detection statistics corresponding to model output for live target data. The malware detection metrics for the live target data can then be characterized using the observed distributions of malware (and malware detections) for the synthetic sample distributions.

Abstract: In a natural language processing model such as a Bidirectional Encoder Representations from Transformers (BERT) model, transformer layers can be replaced with simplified adapters without significant loss of predictive ability. This compressed model may in turn be trained to perform security classification tasks such as detection of new phishing attacks in electronic mail communications.

Abstract: A system for conducting a security recognition task, the system comprising a memory configured to store a model and training data including auxiliary information that will not be available as input to the model when the model is used as a security recognition task model for the security recognition task. The system further comprising one or more processors communicably linked to the memory and comprising a training unit and a prediction unit. The training unit is configured to receive the training data and the model from the memory and subsequently provide the training data to the model, and train the model, as the security recognition task model, using the training data to predict the auxiliary information as well as perform the security recognition task, thereby improving performance of the security recognition task. The prediction unit is configured to use the security recognition task model output to perform the security recognition task while ignoring the auxiliary attributes in the model output.