Generalized Non-deterministic Quantitative Confidence (GNQC) framework

IHIO Innovations Corporation proposes to develop the Generalized Non-deterministic Quantitative Confidence (GNQC) framework that computes novelty and quantifies confidence (or uncertainties) in terms of error bounds in an ML model (algorithm, systems, and agents) for an insufficient data set or never-seen-data. The GNQC framework id based on modified nonparametric statistics to intelligently perform nondeterministic uncertainty quantification. The GNQC enables users to monitor continuously the input data, output and behaviors of machine learning and AI components in terms of error bounds. In Phase I, IHIO will demonstrate the feasibility of GNQC by developing the theoretic basis and testing it on sample scenarios.