The Artificial Intuition™ Platform
Man – Machine thinking together for enhanced decision making
ES – An expert system developed by collaborating subject matter experts focused on diagnostically challenging issues. ES performs like a “System 2” rule-based inferential cognitive process.
ML – Machine learning that employs training data that has been “interpreted” by the ES to drive a predictive analytic. ML performs like a “System 1” experience-based intuitive process in that like humans, intuition is a learned skill that blends prior knowledge with multiple new experiences to establish a new pattern that is immediately recognized.
The Innovation. To be effective in reducing cognitive medical errors in time sensitive critical care settings, we employ a hybrid ES/ML approach that is matched with, and amplifies the System 1/System 2 reasoning processes that expert clinicians use when diagnosing/treating complex patients. The hybrid ES/ML is hosted on our innovative VFusion™ semantic and data science technology stack. Our ES, generated directly by collaborating teams of expert clinicians, is focused on confounding issues/known causes of error, and can perform both as a highly competent real-time “System-2-like” physician diagnostic assistant as well as a ML “big data” training assistant in deriving highly useful natural language processing (NLP) and predictive analytics that can perform as a “System-1-like” assistant.
Our hybrid ES/ML approach enables the exploitation of experience-based heuristic reasoning and nuances machines cannot derive from data alone without expert assistance (e.g. how experts, based on years of experience, interpret patterns in early raw non-specific clinical data and patient context and learn from anomalous patterns). Beyond serving as an exceptionally competent expert system decision aide, our hybrid approach leverages the power of machines to process thousands/millions of cases with the “eyes” of experts (via the ES knowledgebase and NLP).
We believe that beyond ML pattern recognition leading to improved prediction precision, by integrating prior knowledge, our ES-guided machine training may lead to discovery of new novel patterns2 from repositories. We call this highly symbiotic1 real-time cognitive System-1/System-2 assistant “artificial intuition”.
Artificial Intuition Real-Time
- Critical Care Threat Management
- Intelligence, Surveillance and Reconnaissance (ISR) Threat Management
- Tactical Targeting
- Strategic Risk Surveillance
- Insider Threat
1. Licklider JCR. Man-Computer Symbiosis. IRE Trans Hum Factors Electron. 1960;HFE-1(1):4-11. doi:10.1109/THFE2.1960.4503259.
2. Freitas AA. Are we really discovering “interesting” knowledge from data? Expert Updat Spec issue 2nd UK KDD Work. 2006.