Method and apparatus for determining and assessing information to be collected based on information-theoretic measures
First Claim
1. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
- receiving a partial set of information from at least a subset of the information sources for a decision to be made;
performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy;
performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and
when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and
when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and
when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and
when the cost/benefit analysis succeeds, gathering the information from the information source; and
performing a consistency check to determine the validity of the information gathered from the information source; and
when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and
when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and
determining a classification based on the partial set of information;
wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs;
wherein the sensors include sensors selected from a group consisting of seismic sensors, temperature sensors, acoustic sensors, infrared sensors, and radar sensors; and
wherein the sensors are radar sensors, wherein the information gathered is information regarding at least one object, and wherein the method further comprises a step of using the information gathered to determine an object classification for each object about which information was gathered.
1 Assignment
0 Petitions
Accused Products
Abstract
A method, apparatus, and computer program product for determining and assessing information for collection from information sources for a desired level of decision accuracy are presented. Operations include: receiving a partial set of information; performing a minimax entropy-based test to determine a source with useful information; performing a mutual information or a conditional entropy-based test check minimax test validity. With an invalid result, the information source is excluded from further consideration and the minimax test is repeated; with a valid result, a cost/benefit analysis is determines whether to gather the information. If the cost/benefit analysis succeeds, the information is gathered. Otherwise, the information source is excluded and the minimax test is performed again. A consistency check ensures validity of the information prior to restarting the process. Thus, the set of information is iteratively augmented until there is no information to add or until adding information would be cost-prohibitive.
-
Citations
96 Claims
-
1. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; wherein the sensors include sensors selected from a group consisting of seismic sensors, temperature sensors, acoustic sensors, infrared sensors, and radar sensors; and wherein the sensors are radar sensors, wherein the information gathered is information regarding at least one object, and wherein the method further comprises a step of using the information gathered to determine an object classification for each object about which information was gathered. - View Dependent Claims (2, 3)
-
-
4. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and wherein the sensors provide information regarding routing of network communications in a computer network, and wherein the method further comprises a step of using the information gathered to determine optimal paths for routing signals within the computer network. - View Dependent Claims (5)
-
-
6. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and wherein the information sources provide information for optimizing an investment portfolio, and wherein the method further comprises a step of using the information gathered to provide a recommendation regarding an investment to be made. - View Dependent Claims (7)
-
-
8. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and wherein the sensors provide information regarding processing capabilities of a plurality of data processors in a distributed computing environment and about current processing jobs within the plurality of data processors, and wherein the method further comprises a step of using the information gathered to determine the processors to use for optimal processing of a new processing job.
-
-
9. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and wherein the information sources are selected from a group consisting of databases of marketing information and market studies to be performed by a user, and wherein the method further comprises a step of using the information gathered to make marketing-related recommendations.
-
-
10. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and wherein the information sources are medical information sources, and wherein the method further comprises a step of using the medical information gathered to diagnose a medical patient or to recommend a treatment for a medical patient. - View Dependent Claims (11, 12, 13)
-
-
14. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and wherein the sensors monitor manufacturing activities, and wherein the method further comprises a step of using the information gathered to allocate resources in a manufacturing process in order to optimize the use of resources in the manufacturing process.
-
-
15. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the steD of performing the minimax entropy-based test on the auamented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and wherein the information sources are terrorist-related data sources, and wherein the method further comprises a step of using the information gathered to track potential terrorists.
-
-
16. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and wherein the sensors provide feedback from individual ones of a plurality of engines on a vehicle, and wherein the method further comprises a step of determining which engine or combination of engines to use for a particular maneuver of the vehicle.
-
-
17. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a group consisting of databases, sensors, and user inputs; and wherein the information sources provide information regarding resource allocation on a battlefield, and wherein the method further comprises a step of using the information gathered to make recommendations regarding future resource allocation on the battlefield in order to optimize the use of resources on the battlefield.
-
-
18. A method for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the method comprising steps of:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information; wherein the information sources include a plurality of information sources selected from a aroup consisting of databases, sensors, and user inputs; and where the user input is information to aid in solving a computer problem, and wherein the method further comprises a step of using the information gathered to make recommendations regarding the most likely way to correct the problem.
-
-
19. A computer program product for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy the computer program product comprising means, stored on a computer readable medium, for:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a mutual information entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the means for performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information. - View Dependent Claims (20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44)
-
-
45. An apparatus for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the apparatus comprising a computer system including a processor, a memory coupled with the processor, an input coupled with the processor for receiving the information from the information sources, and an output coupled with the processor for outputting an output selected from a group consisting of a recommendation, a decision, and a classification based on the information collected, wherein the computer system further comprises means, residing in its processor and memory, for:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information. - View Dependent Claims (46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70)
-
-
71. An apparatus for determining and assessing information to be collected from a set of information sources for a desired level of decision accuracy, the apparatus comprising a plurality information sources and a computer system including a processor, a memory coupled with the processor, an input communicatively coupled with the information source and with the processor for receiving information from the information sources, and an output coupled with the processor for outputting an output selected from a group consisting of a recommendation, a decision, and a classification based on the information collected, wherein the computer system further comprises means, residing in its processor and memory, for:
-
receiving a partial set of information from at least a subset of the information sources for a decision to be made; performing a minimax entropy-based test to determine an information source from which to next gather information that is most likely to improve the decision accuracy; performing a check based on a test selected from a group consisting of a mutual information entropy-based test and a conditional entropy-based test to determine whether the minimax entropy-based test generated a valid or an invalid result; and when the minimax entropy-based test is determined to have generated an invalid result, excluding the information source from the set of information sources to produce a set of remaining information sources, and repeating the minimax entropy-based test on the remaining information sources; and when the minimax entropy-based test is determined to have generated a valid result, performing a cost/benefit analysis to determine whether the information should be gathered from the information source; and when the cost/benefit analysis fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; and when the cost/benefit analysis succeeds, gathering the information from the information source; and performing a consistency check to determine the validity of the information gathered from the information source; and when the consistency check fails, excluding the information source from further consideration and repeating the minimax entropy-based test on the remaining information sources; when the consistency check succeeds, gathering the information from the information source and adding the information to the partial set of information to produce an augmented partial set of information, and starting again with the step of performing the minimax entropy-based test on the augmented partial set of information, whereby the partial set of information is iteratively augmented with additional information until there is no further information to add or until adding further information would be cost-prohibitive; and determining a classification based on the partial set of information. - View Dependent Claims (72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96)
-
Specification