Method for making optimal selections based on multiple objective and subjective criteria

• US 8,886,586 B2
• Filed: 05/18/2010
• Issued: 11/11/2014
• Est. Priority Date: 05/24/2009
• Status: Expired due to Fees

• Current Assignee: Pi-Coral Incorporated
• Original Assignee: Pi-Coral Incorporated
• Inventors: Osmond, Roger Frederick
• Primary Examiner(s): Gaffin, Jeffrey A
• Assistant Examiner(s): Sugent, James F
• Application Number:
  US12/782,495
• Publication Number:
  US 20100299298A1
• Time in Patent Office: 1,638 Days
• Field of Search: 706/20, 706/25, 706/46, 706/47
• US Class Current: 706/47
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/008 based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G06N 5/025 Extracting rules from data
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

A method for making optimal decisions includes the ability to consider and weigh multiple factors, where those factors might be numeric, or non-numeric, objective or subjective. Further, the method ensures that factors are not prematurely eliminated. This contrasts with behavior common to decision-tree based approaches. The method further allows for weighting based on multiple statistical means as well as by the application of non-statistical values.

Claims

1. A computer-implemented method for making optimal decisions, the method comprising, by a processor:

receiving a plurality of candidates from an interested party;

receiving a plurality of criteria associated with a characteristic of the plurality of candidates, each of the plurality of criteria being associated with a type;

receiving values, from the interested party, for each of the plurality of criteria for each of the plurality of candidates, wherein the values have one of a numerical form or a non-numerical form based on the type;

normalizing the values for each of the plurality of criteria by applying normalization rules associated with each of the plurality of criteria, wherein the normalization rules comprise at least one of minimum and maximum values, expected distribution, and range of values;

generating a raw candidate ranking by ranking the plurality of candidates for each of the plurality of criteria;

receiving a context-based weight model for each of the plurality of criteria, each context-based weight model being configured to indicate a priority of each of the plurality of criteria wherein the context-based weight model comprises an exponential weighting rule configured to prioritize a most important criteria of the interested party;

generating a weighted ranking of the plurality of candidates by applying the context-based weight model for each of the plurality of criteria to the raw candidate ranking; and

sorting the plurality of candidates based on the weighted ranking, thereby generating a weighted list of the plurality of candidates sorted through a comparison of each of the plurality of criteria without premature candidate elimination.

2. The computer-implemented method of claim 1, wherein each candidate is uniquely identifiable and represents a decision option.

3. The computer-implemented method of claim 1, wherein the values represent a relationship between the plurality of candidates and the plurality of criteria.

4. The computer-implemented method of claim 1, wherein the values comprise subjective values.

5. The computer-implemented method of claim 1, wherein the values comprise at least one of the following:

subjective values, objective values and context-based values.

6. The computer-implemented method of claim 1, wherein the values are normalized using at least one of the following normalization rules:

existential, proportional, stepped, deviation, and logarithmic.

7. The computer-implemented method of claim 6, wherein the normalization rules are applied on a per-criterion basis.

8. The computer-implemented method of claim 1, wherein at least a portion of the plurality of criteria are associated with values comprising pre-ordered sets of values.

9. The computer-implemented method of claim 8, wherein the pre-ordered sets of values are pre-normalized according to at least one pre-defined rule.

10. The computer-implemented method of claim 1, wherein the weight model comprises at least one of the following models:

exponential, raw, linear, logarithmic, square, and stepped.