Data-Intensive Computing Initiative (DICI)
Research Area: Analytic Algorithms
Analytical processes are necessary to perform assessments of the data, to identify patterns and anomalies, and to provide feature identification and classification. One of the key outcomes of the DICI is to perform these analysis processes in situ to the data stream. This requires the ability to analyze data as it is collected to perform analysis in real-time. By developing techniques to perform real-time analyses on data before it is stored, we can reduce data transfer size and provide an intermediate product that expedites the time-to-solution for critical processes that require immediate decisions. Additionally, by assessing the output of these algorithms, the workflow for a specific problem may be dynamically adapted.
DICI projects that support this R&D capability:
- High Performance Data Analysis Tools for Intelligent Mass Spectrometry Pipelines
- CLIQUE: Correlation Layers for Information Query and Exploration
- Adaptive Composite Analysis for Complex Systems
- Data-Intensive Machine Learning for Real-Time Decision Analysis
Contact: Paul Whitney