Mark Temple-Raston, LLC, USA

Scientific Machine Learning implements the science-of-counting to analytically process any time-series and produce a complete set of thermodynamic measurements that define the state of the system. The scientific measurements are illustrated with a time-series of closing-prices for a stock (GE). Exact scientific measurements from Scientific Machine Learning (SML) are then used to create time-series decision services without model or bias. Services are built for a large class of Open allocation problems and applied to real optimal sales data for a consumer product good.

Machine Learning, Risk Analysis, Non-Equilibrium Thermodynamics, Information Theory, Decision Theory