Morpheus Models

Morpheus comes with a number of pretrained models with corresponding training, validation scripts, and datasets. The latest release of these models can be found here.


GPU Mem Req


Anomalous Behavior Profiling (ABP)


This model is an example of a binary classifier to differentiate between anomalous GPU behavior such as crypto mining / GPU malware, and non-anomalous GPU-based workflows (for example, ML/DL training). The model is an XGBoost model.

Digital Fingerprinting (DFP)


This use case is currently implemented to detect changes in a users’ behavior that indicates a change from a human to a machine or a machine to a human. The model is an ensemble of an Autoencoder and fast Fourier transform reconstruction.

Fraud Detection


This model shows an application of a graph neural network for fraud detection in a credit card transaction graph. A transaction dataset that includes three types of nodes, transaction, client, and merchant nodes is used for modeling. A combination of GraphSAGE along with XGBoost is used to identify frauds in the transaction networks.

Ransomware Detection Model


This model shows an application of DOCA AppShield to use data from volatile memory to classify processes as ransomware or bengin. This model uses a sliding window over time and feeds derived data into a random forest classifiers of various lengths depending on the amount of data collected.

Flexible Log Parsing


This model is an example of using Named Entity Recognition (NER) for log parsing, specifically Apache HTTP Server logs.

© Copyright 2023, NVIDIA. Last updated on Apr 11, 2023.