State inference in a heterogeneous system

An apparatus for inferring a state of a system of interest, said apparatus comprising:

• at least one processor and memory comprising computer program code, the memory and the computer program code configured to, with the at least one processor, cause the apparatus to:
• determine a first measurement value for a first indicator and a second measurement value for a second indicator in the system of interest, where the first measurement value and the second measurement value are indicative of the state of the system of interest;
• retrieve first database values for the first indicator and second database values for the second indicator, wherein the first database values include values of at least one example of a system that is known to be in a control state and at least one example of a system that is known to be in a comparison state and wherein the second database values include values of at least one example of a system that is known to be in a control state and at least one example of a system that is known to be in a comparison state;
• determine a measure of goodness for said first indicator by applying a statistical test to said first database values and determine a measure of goodness for said second indicator by applying a statistical test to said second database values, said measure of goodness for said first indicator and said measure of goodness for said second indicator indicating how significant said first indicator and said second indicator are statistically in differentiating between said control state and said comparison state;
• determine a difference value for said first measurement value and determine a difference value for said second measurement value, said difference value for said first measurement value and said difference value for said second measurement value indicating a relative distance of the system of interest from said control state when compared with said comparison state;
• define a composite difference value describing how closely the first measurement value and the second measurement value match with the corresponding database values of at least one of the comparison state or the control state;
• determine a composite measure of goodness for the composite difference value by using information indicative of said measure of goodness of said first indicator and said measure of goodness for said second indicator, said composite measure of goodness for the composite difference value indicating how significant said composite difference value is statistically in differentiating between said control state and said comparison state;
• form an initial tree structure comprising a plurality of nodes, said plurality of nodes being indicative of said first and second measurement values and one or more of the first and second difference values and being based upon at least said composite difference value and said composite measure of goodness;
• determine a total difference value for the system of interest, said total difference value indicating a relative distance of the system of interest from said control state when compared with said comparison state;
• re-order one or more of said plurality of nodes in said initial tree structure based on the measures of goodness or the difference values, said re-ordering resulting in a reordered tree structure;
• generate a visualization of at least said reordered tree structure for a user of said apparatus for inferring a state of a system of interest, said visualization being accessible to said user via a user interface;
• and
• dynamically refine said reordered tree structure said dynamic refining occurring in response to one or more user selections related to said first and second measurement values and one or more difference values.

A system for inferring a state of a system of interest, comprising at least one processor, memory including computer program code, the memory and the computer program code configured to, with the at least one processor, cause the system to:

• determine a first measurement value for a first indicator and a second measurement value for a second indicator in the system of interest, where the first measurement value and the second measurement value are indicative of the state of the system of interest;
• retrieve first database values for the first indicator and second database values for the second indicator, wherein the first database values include values of at least one example of a system that is known to be in a control state and at least one example of a system that is known to be in a comparison state and wherein the second database values include values of at least one example of a system that is known to be in a control state and at least one example of a system that is known to be in a comparison state;
• determine a measure of goodness for said first indicator by applying a statistical test to said first database values and determine a measure of goodness for said second indicator by applying a statistical test to said second database values, said measure of goodness for said first indicator and said measure of goodness for said second indicator indicating how significant said first indicator and said second indicator are statistically in differentiating between said control state and said comparison state;
• determine a difference value for said first measurement value and determine a difference value for said second measurement value, said difference value for said first measurement value and said difference value for said second measurement value indicating a relative distance of the system of interest from said control state when compared with said comparison state;
• define a composite difference value describing how closely the first measurement value and the second measurement value match with the corresponding database values of at least one of the comparison state or the control state;
• determine a composite measure of goodness for the composite difference value by using information indicative of said measure of goodness of said first indicator and said measure of goodness for said second indicator, said composite measure of goodness for the composite difference value indicating how significant said composite difference value is statistically in differentiating between said control state and said comparison state;
• form an initial tree structure comprising a plurality of nodes, said plurality of nodes being indicative of said first and second measurement values and one or more of the first and second difference values and being based upon at least said composite difference value and said composite measure of goodness;
• determine a total difference value for the system of interest, said total difference value indicating a relative distance of the system of interest from said control state when compared with said comparison state;
• re-order one or more of said plurality of nodes in said initial tree structure based on the measures of goodness or the difference values, said re-ordering resulting in a reordered tree structure;
• generate a visualization of at least said reordered tree structure for a user of said apparatus for inferring a state of a system of interest, said visualization being accessible to said user via a user interface;
• and
• dynamically refine said reordered tree structure said dynamic refining occurring in response to one or more user selections related to said first and second measurement values and one or more difference values.

A computer program product for inferring a state of a system of interest, said computer program product being stored on a non-transitory computer readable medium and executable in a data processing device, the computer program product comprising one or more computer program code sections configured for:

• determining a first measurement value for a first indicator and a second measurement value for a second indicator from the system of interest, where the first measurement value and the second measurement value are indicative of the state of the system of interest,
• retrieving first database values for the first indicator and second database values for the second indicator, wherein the first database values include values of at least one example of a system that is known to be in a control state and at least one example of a system that is known to be in a comparison state and wherein the second database values include values of at least one example of a system that is known to be in a control state and at least one example of a system that is known to be in a comparison state;
• determining a measure of goodness for said first indicator by applying a statistical test to said first database values and determining a measure of goodness for said second indicator by applying a statistical test to said second database values, said measure of goodness for said first indicator and said measure of goodness for said second indicator indicating how significant said first indicator and said second indicator are statistically in differentiating between said control state and said comparison state,
• determining a difference value for said first measurement value and determining a difference value for said second measurement value, said difference value for said first measurement value and said difference value for said second measurement value indicating a relative distance of the system of interest from said control state when compared with said comparison state,
• defining a composite difference value describing how closely the first measurement value and the second measurement value match with the corresponding database values of the comparison state and/or the control state,
• determining a composite measure of goodness for the composite difference value by using information indicative of said measure of goodness for said first indicator and said measure of goodness for said second indicator, said composite measure of goodness for the composite difference value indicating how significant said composite difference value is statistically in differentiating between said control state and said comparison state,
• forming an initial tree structure comprising a plurality of nodes, said plurality of nodes being indicative of said first and second measurement values and one or more of the first and second difference values and being based upon at least said composite difference value and said composite measure of goodness,
• determining a total difference value for the system of interest, said total difference value indicating a relative distance of the system of interest from said control state when compared with said comparison state,
• re-ordering one or more of said plurality of nodes in said initial tree structure based on the measures of goodness or the difference values, said re-ordering resulting in a reordered tree structure;
• generating a visualization of at least said reordered tree structure for a user of said apparatus for inferring a state of a system of interest, said visualization being accessible to said user via a user interface;
• and
• dynamically refining said reordered tree structure said dynamic refining occurring in response to one or more user selections related to said first and second measurement values and one or more difference values.