Multi-scale information dynamics for decision making

A computer-implemented method for automated discovery of unknown patterns from multiple heterogeneous sensors and determining value of new data acquisitions, comprising:

• an act of causing a data processor to execute instructions stored on a non-transitory memory such that upon execution, the data processor performs operations of:
• processing input data from a plurality of surveillance sensors associated with a plurality of unmanned aerial vehicles with a coarse-graining module, wherein the input data is processed with a plurality of time windows and time delays to generate multi-scale symbolic representations of the input data having various levels of granularity;
• discovering and extracting previously unknown patterns in the multi-scale symbolic representations with a local information dynamics module;
• determining, for each data point in time or for each surveillance sensor, a value of acquiring new input data at a next data point in time or from a new surveillance sensor with a non-myopic value of information module, resulting in a plurality of values of acquiring new input data;
• generating a ranking of the plurality of values of acquiring new input data with the non-myopic value of information module;
• and
• using the ranking, selecting at least one of the plurality of surveillance sensors associated with an unmanned aerial vehicle from which to collect new input data.

A computer program product for automated discovery of unknown patterns from multiple heterogeneous sensors and determining value of new data acquisitions, the computer program product comprising computer-readable instructions stored on a non-transitory computer-readable medium that are executable by a computer having a processor for causing the processor to perform operations of:

• processing input data from a plurality of surveillance sensors associated with a plurality of unmanned aerial vehicles with a coarse-graining module, wherein the input data is processed with a plurality of time windows and time delays to generate multi-scale symbolic representations of the input data having various levels of granularity;
• discovering and extracting previously unknown patterns in the multi-scale symbolic representations with a local information dynamics module;
• determining, for each data point in time or for each surveillance sensor, a value of acquiring new input data at a next data point in time or from a new surveillance sensor with a non-myopic value of information module, resulting in a plurality of values of acquiring new input data;
• generating a ranking of the plurality of values of acquiring new input data with the non-myopic value of information module;
• and
• using the ranking, selecting at least one of the plurality of surveillance sensors associated with an unmanned aerial vehicle from which to collect new input data.