HOW IT WORKS
STEP ONE
The Ingestion Process (“Human on the Loop”)
- Making the data the best it can be
- Curating raw data using Machine Learning to translate it to a common ontology
- We can ingest data from virtually any source.
- Our capacity is completely elastic and capable of updating in real-time. That means social media, government statistics, polling, public opinion, or anything else. This allows you to see the effect of modeling as it happens, not days later.
- We understand that sources, timing, and reliability of data are the foundation of reliable forecasting.
STEP TWO
Building the Neural Network
The input of curated data (“Human on the Loop”)
The curated data workflow is designed to reduce the statistical biases common to AI/ML systems. The output is a neural network of relationships where small subgroups remain rather than being lost as statistical noise. The resulting synthetic model includes all data relationships so all queries have already been answered.
RWISE then begins to build the neural connections that are the foundation for the modeling of complex impacts of events. Unlike virtually all other platforms, RWISE integrates multiple technologies in the neural development phase. This makes data a more global in its use during simulations. Unlike others, we can capture the impact of different perspectives.
STEP THREE
Agent-Based Modeling (“Human in the Loop”)
- Determining the coefficient baseline of the future
- Apply Variable Conditions and Changes
- View Impact on Future Performance/Results
