Measuring sensitivity of a factor in a decision

• US 8,732,115 B1
• Filed: 02/11/2013
• Issued: 05/20/2014
• Est. Priority Date: 12/23/2009
• Status: Expired due to Fees

• Current Assignee: Decision Lens, Inc.
• Original Assignee: Decision Lens, Inc.
• Inventors: Adams, William James Louis, Saaty, Daniel Lowell
• Primary Examiner(s): Rifkin, Ben M
• Assistant Examiner(s): Seck, Ababacar
• Application Number:
  US13/764,010
• Time in Patent Office: 463 Days
• Field of Search: 706/55
• US Class Current: 706/55
• CPC Class Codes: G06N 3/02 Neural networks

Abstract

An analytic network process (ANP) storage memory stores an ANP weighted supermatrix representing an ANP model. A processor is in communication with the ANP storage memory. The processor is configured to change priorities of a node in the ANP weighted supermatrix to be more important, to change priorities of the node in the ANP weighted supermatrix to be less important, and to assess a sensitivity of the node which was changed relative to the ANP model. The processor further is configured to maintain a same proportionality in the ANP weighted supermatrix for the changing of the priorities and the assessing of the relative sensitivity.

Claims

1. An apparatus comprising:

an analytic network process (ANP) storage memory that stores an ANP weighted supermatrix representing an ANP model; and

a processor in communication with the ANP storage memory, the processor being configured to facilitate

changing priorities of a node in the ANP weighted supermatrix to be more important, and changing priorities of the node in the ANP weighted supermatrix to be less important;

assessing a sensitivity of the node itself which was changed relative to the ANP model; and

maintaining a same proportionality in the ANP weighted supermatrix for the changing of the priorities and the assessing of the relative sensitivity,

wherein the sensitivity of the node itself is assessed by changing, by a same predetermined amount, each entry throughout a same row of the node of the ANP weighted supermatrix and consequently rescaling the rest of the entries in a column of the ANP weighted supermatrix that corresponds to the node being assessed,

the changing comprising, for all 0≦

p≦

1, defining F_W,r,p0(p) by changing the r^th row and then rescaling the remaining entries in the columns so that the columns continue to add to one;

rescaling to be less important comprising, for 0≦

p≦

p₀, changing the r^th row by scaling it by

2. The apparatus of claim 1, the priorities of the node being changed to be more important by increasing the priorities toward 1, and scaling entries in other columns that do not include the node.

3. The apparatus of claim 1, the priorities of the node being changed to be less important by decreasing the priorities toward 0, scaling a row corresponding to the node, and renormalizing the columns.

5. The apparatus of claim 1, further comprising:

an output unit in communication with the processor and configured to output, for a display to a user, the relative sensitivity of the node which was changed.

6. The apparatus of claim 1, further comprising

an input unit configured to input, from an input device, pairwise comparisons, ANP ratings, or ANP client data, which are transformed into priority vectors and stored into the ANP weighted supermatrix, the pairwise comparisons representing a judgment of priority between ANP alternatives in the pair, the ANP ratings representing a rating of a choice, and the ANP client data representing real world values.

4. An apparatus comprising:

an analytic network process (ANP) storage memory that stores an ANP weighted supermatrix representing an ANP model; and

a processor in communication with the ANP storage memory, the processor being configured to facilitate

changing priorities of a node in the ANP weighted supermatrix to be more important, and changing priorities of the node in the ANP weighted supermatrix to be less important;

assessing a sensitivity of the node itself which was changed relative to the ANP model; and

maintaining a same proportionality in the ANP weighted supermatrix for the changing of the priorities and the assessing of the relative sensitivity,

wherein the sensitivity of the node itself is assessed by changing, by a same predetermined amount, each entry throughout a same row of the node of the ANP weighted supermatrix and consequently rescaling the rest of the entries in a column of the ANP weighted supermatrix that corresponds to the node being assessed,

the changing to be less important comprising, when 0≦

p≦

p₀, defining F_W,r,p0(p) by scaling the r^th row by

7. A method, comprising:

storing, in an analytic network process (ANP) storage memory, an ANP weighted supermatrix representing an ANP model;

in a processor in communication with the ANP storage memory, changing priorities of a node in the ANP weighted supermatrix to be more important;

assessing a sensitivity of the node itself which was changed relative to the ANP model; and

maintaining a same proportionality in the ANP weighted supermatrix for the changing of the priorities and the assessing of the relative sensitivity,

wherein the sensitivity of the node itself is assessed by changing, by a same predetermined amount, each entry throughout a same row of the node of the ANP weighted supermatrix and consequently rescaling the rest of the entries in a column of the ANP weighted supermatrix that corresponds to the node being assessed,

the changing comprising, for all 0≦

p≦

1, defining F_W,r,p0(p) by changing the r^th row and then rescaling the remaining entries in the columns so that the columns continue to add to one;

rescaling to be more important comprising, for p₀≦

p≦

1, changing the entries in the r^th row consistent with the following formula

f(p)=F_W,r,p0(p)_r,j=1−

α

(1−

W_r,j)

where

8. The method of claim 7, the priorities of the node being changed to be more important by increasing the priorities toward 1, and scaling entries in other columns that do not include the node.

9. The method of claim 7, further comprising changing priorities of the node in the ANP weighted supermatrix to be less important, the priorities of the node being changed to be less important by decreasing the priorities toward 0, scaling a row corresponding to the node, and renormalizing the columns.

11. The method of claim 7, further comprising:

outputting, to an output unit in communication with the processor, the relative sensitivity of the node which was changed.

12. The method of claim 7, further comprising

inputting, from an input device, pairwise comparisons, ANP ratings, or ANP client data, which are transformed into priority vectors and stored into the ANP weighted supermatrix, the pairwise comparisons representing a judgment of priority between ANP alternatives in the pair, the ANP ratings representing a rating of a choice, and the ANP client data representing real world values.

10. A method, comprising:

storing, in an analytic network process (ANP) storage memory, an ANP weighted supermatrix representing an ANP model;

in a processor in communication with the ANP storage memory, changing priorities of a node in the ANP weighted supermatrix to be more important;

assessing a sensitivity of the node itself which was changed relative to the ANP model; and

maintaining a same proportionality in the ANP weighted supermatrix for the changing of the priorities and the assessing of the relative sensitivity,

wherein the sensitivity of the node itself is assessed by changing, by a same predetermined amount, each entry throughout a same row of the node of the ANP weighted supermatrix and consequently rescaling the rest of the entries in a column of the ANP weighted supermatrix that corresponds to the node being assessed,

the changing to be more important comprising, when p₀≦

p≦

1, defining F_W,r,p0(p) by leaving alone columns of W for which W_r,i=0 and scaling all entries in the other columns, except for the entry in the r^th row, by

13. A non-transitory computer-readable storage medium encoded with a computer executable instructions, wherein execution of said computer executable instructions by one or more processors causes a computer to perform the steps of:

storing, in an analytic network process (ANP) storage memory, an ANP weighted supermatrix representing an ANP model;

changing priorities of a node in the ANP weighted supermatrix to be less important;

assessing a sensitivity of the node itself which was changed relative to the ANP model; and

maintaining a same proportionality in the ANP weighted supermatrix for the changing of the priorities and the assessing of the relative sensitivity,

wherein the sensitivity of the node itself is assessed by changing, by a same predetermined amount, each entry throughout a same row of the node of the ANP weighted supermatrix and consequently rescaling the rest of the entries in a column of the ANP weighted supermatrix that corresponds to the node being assessed,

further comprising changing the priorities of the node to be more important,

the changing comprising, for all 0≦

p≦

1, defining F_W,r,p0(p) by changing the r^th row and then rescaling the remaining entries in the columns so that the columns continue to add to one;

rescaling to be less important comprising, for 0≦

p≦

p₀, changing the r^th row by scaling it by

14. The computer readable storage medium of claim 13, further comprising changing the priorities of the node to be more important, the priorities of the node being changed to be more important by increasing the priorities toward 1, and scaling entries in other columns that do not include the node.

15. The computer readable storage medium of claim 13, the priorities of the node being changed to be less important by decreasing the priorities toward 0, scaling a row corresponding to the node, and renormalizing the columns.

17. The computer readable storage medium of claim 13, further comprising instructions for:

outputting, to an output unit in communication with the processor, the relative sensitivity of the node which was changed.

18. The computer readable storage medium of claim 13, further comprising instructions for:

inputting, from an input device, pairwise comparisons, ANP ratings, or ANP client data, which are transformed into priority vectors and stored into the ANP weighted supermatrix, the pairwise comparisons representing a judgment of priority between ANP alternatives in the pair, the ANP ratings representing a rating of a choice, and the ANP client data representing real world values.

16. A non-transitory computer-readable storage medium encoded with a computer executable instructions, wherein execution of said computer executable instructions by one or more processors causes a computer to perform the steps of:

storing, in an analytic network process (ANP) storage memory, an ANP weighted supermatrix representing an ANP model;

changing priorities of a node in the ANP weighted supermatrix to be less important;

assessing a sensitivity of the node itself which was changed relative to the ANP model; and

maintaining a same proportionality in the ANP weighted supermatrix for the changing of the priorities and the assessing of the relative sensitivity,

wherein the sensitivity of the node itself is assessed by changing, by a same predetermined amount, each entry throughout a same row of the node of the ANP weighted supermatrix and consequently rescaling the rest of the entries in a column of the ANP weighted supermatrix that corresponds to the node being assessed,

the changing to be less important comprising, when 0≦

p≦

p₀, defining F_W,r,p0(p) by scaling the r^th row by