Predictive Analysis of ISR Track Data Using Decentralized Machine Learning (DML)

Prescient Edge proposes building upon our commercially deployed Decentralized Machine Learning platform to develop a real-time intelligence analysis tool for data sharing the Joint All-Domain environment. Decentralized Machine Learning (DML) is a group of machine learning approaches that allow neural networks to train on distributed data without requiring expensive data storage centers or high bandwidth to transmit complex metadata. Prescient Edge’s DML software is currently supporting the healthcare industry to enable AI models to be trained in a distributed network of sensitive healthcare data without centralized data storage. Due to similarities in privacy and security requirements between DoD and the Healthcare sector (such as GDPR and HIPPA regulations), a decentralized machine learning approach is the optimal choice for training predictive analysis tools in a secure environment. For this Phase I effort, Prescient Edge will conduct a feasibility study for models, algorithms, and network architectures that can deliver pattern of life analysis and anomaly detection in support of Air Force intel and other intelligence community partners.