US6792399B1
Expired / Lapsed
US
⏱️ 0.0 years remaining
Combination forecasting using clusterization
Assignee: C4CASTCOM Inc
•
Filing: 1999-09-08
•
Granted: 2004-09-14
•
Expires: 2019-09-08
Abstract
Abstract Combination forecasts are generated using predictions obtained from a group of forecasters. The forecasters are first divided into a number of pre-defined clusters, which typically will have been formed using statistical clustering techniques. In particular, clusters of forecasters can be formed based on similarities of the forecasters' predictions. Then, statistical data are calculated for each pre-defined cluster (e.g., measures of central tendency and dispersion). Finally, the statistical data for all the pre-defined clusters are combined so as to obtain a combination forecast.
Inventors
- 👤 G. Michael Phillips
- 👤 William P. Jennings
- 👤 M. Chapman Findlay, III
- 👤 Stephen A. Klein
- 👤 Mark E. Rice
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 17:05
CPC Classifications
G; G06; G06Q; G06Q10/00; G06Q10/06; G06Q40/00; G06Q40/04; G06Q40/06
Claims Analysis
Total Claims: 48
Claim 1
Independent Claim
93 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; and combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters were defined based on a comparison of previous predictions made by the forecasters.
Claim 2
Independent Claim
93 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; and combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters were defined based on a comparison of previous predictions made by the forecasters.
Claim 3
Dependent Claim
18 words
A method according to claim 1 , wherein the clusters were defined by utilizing a statistical clustering technique.
Claim 4
Dependent Claim
29 words
A method according to claim 1 , wherein the combination forecast is a forecast of a value of a variable that has been predicted by the group of forecasters.
Claim 5
Dependent Claim
28 words
A method according to claim 3 , further comprising the step of utilizing the combination forecast in an interpolation model to predict the value of a different variable.
Claim 6
Dependent Claim
16 words
A method according to claim 1 , wherein the forecasters comprise a large longitudinal forecast panel.
Claim 7
Dependent Claim
14 words
A method according to claim 1 , wherein the forecasters comprise thousands of people.
Claim 8
Dependent Claim
20 words
A method according to claim 1 , wherein the clusters are formed simultaneously with assigning the forecasters to said clusters.
Claim 9
Dependent Claim
103 words
A computer-readable medium storing computer-executable process steps for providing combination forecasts, said process steps comprising steps to: obtain predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; divide the forecasters into plural clusters by utilizing a statistical clustering technique; calculate statistical data for each cluster; and combine the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters were defined based on a comparison of previous predictions made by the forecasters.
Claim 10
Dependent Claim
103 words
A computer-readable medium storing computer-executable process steps for providing combination forecasts, said process steps comprising steps to: obtain predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; divide the forecasters into plural clusters by utilizing a statistical clustering technique; calculate statistical data for each cluster; and combine the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters were defined based on a comparison of previous predictions made by the forecasters.
Claim 11
Independent Claim
107 words
An apparatus for providing combination forecasts, said apparatus comprising: (a) means for obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; (b) means for dividing the forecasters into plural clusters by utilizing a statistical clustering technique; (c) means for calculating statistical data for each cluster; and (d) means for combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters were defined based on a comparison of previous predictions made by the forecasters.
Claim 12
Independent Claim
107 words
An apparatus for providing combination forecasts, said apparatus comprising: (a) means for obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; (b) means for dividing the forecasters into plural clusters by utilizing a statistical clustering technique; (c) means for calculating statistical data for each cluster; and (d) means for combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters were defined based on a comparison of previous predictions made by the forecasters.
Claim 13
Dependent Claim
16 words
An apparatus according to claim 9 , wherein the forecasters comprise a large longitudinal forecast panel.
Claim 14
Dependent Claim
14 words
An apparatus according to claim 9 , wherein the forecasters comprise thousands of people.
Claim 15
Dependent Claim
20 words
An apparatus according to claim 9 , wherein the clusters are formed simultaneously with assigning the forecasters to said clusters.
Claim 16
Independent Claim
110 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and repeating said obtaining, dividing, calculating and combining steps for each of plural different prediction events, wherein a different subset of the group of forecasters participates in each of the plural different prediction events.
Claim 17
Independent Claim
110 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and repeating said obtaining, dividing, calculating and combining steps for each of plural different prediction events, wherein a different subset of the group of forecasters participates in each of the plural different prediction events.
Claim 18
Independent Claim
94 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; and combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the statistical data calculated for each cluster includes a central tendency and a measure of dispersion.
Claim 19
Independent Claim
94 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; and combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the statistical data calculated for each cluster includes a central tendency and a measure of dispersion.
Claim 20
Dependent Claim
36 words
A method according to claim 14 , wherein when the statistical data for all the clusters are combined, the central tendency for each cluster is weighted according to the measure of dispersion for said each cluster.
Claim 21
Dependent Claim
16 words
A method according to claim 14 , wherein the central tendency comprises a historical trimmed median.
Claim 22
Dependent Claim
37 words
A method according to claim 14 , wherein when the statistical data for all the clusters are combined, the central tendency for each cluster is weighted according to previous prediction accuracy of forecasters within said each cluster.
Claim 23
Dependent Claim
31 words
A method according to claim 17 , wherein the central tendency for each cluster is weighted according to prediction accuracy of central tendencies of previous predictions originating from said each cluster.
Claim 24
Independent Claim
88 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; and combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters have been defined using a statistical regression technique.
Claim 25
Independent Claim
88 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; and combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters have been defined using a statistical regression technique.
Claim 26
Independent Claim
97 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and identifying an assignment formula for assigning a new forecaster to a cluster based on personal characteristics of said new forecaster.
Claim 27
Independent Claim
97 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and identifying an assignment formula for assigning a new forecaster to a cluster based on personal characteristics of said new forecaster.
Claim 28
Dependent Claim
33 words
A method according to claim 20 , wherein said assignment formula minimizes a distance based on said personal characteristics between the new forecaster and the cluster to which the new forecaster is assigned.
Claim 29
Dependent Claim
41 words
A method according to claim 20 , wherein said dividing step comprises assigning a first portion of the forecasters to clusters based on previous prediction similarities and assigning a second portion of the forecasters to clusters based on the assignment formula.
Claim 30
Independent Claim
99 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; and combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters are defined based on a comparison of previous predictions made by the forecasters for at least two different variables.
Claim 31
Independent Claim
99 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; and combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure, wherein the clusters are defined based on a comparison of previous predictions made by the forecasters for at least two different variables.
Claim 32
Independent Claim
88 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and periodically re-clustering the forecasters to obtain a new set of clusters.
Claim 33
Independent Claim
88 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and periodically re-clustering the forecasters to obtain a new set of clusters.
Claim 34
Independent Claim
109 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into a first set of clusters by utilizing a statistical clustering technique based on previous predictions for the forecasters; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; re-clustering the forecasters to obtain a second set of clusters; and utilizing the second set of clusters to generate an additional combination forecast.
Claim 35
Independent Claim
109 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into a first set of clusters by utilizing a statistical clustering technique based on previous predictions for the forecasters; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; re-clustering the forecasters to obtain a second set of clusters; and utilizing the second set of clusters to generate an additional combination forecast.
Claim 36
Dependent Claim
45 words
A method according to claim 25 , further comprising a step of repeating said obtaining, dividing, calculating and combining steps for each of plural different prediction events, wherein a different subset of the group of forecasters participates in each of the plural different prediction events.
Claim 37
Dependent Claim
23 words
A method according to claim 25 , wherein the statistical data calculated for each cluster in the first set includes a central tendency.
Claim 38
Dependent Claim
25 words
A method according to claim 27 , wherein the statistical data calculated for each cluster in the first set also includes a measure of dispersion.
Claim 39
Dependent Claim
36 words
A method according to claim 25 , further comprising a step of identifying an assignment formula for assigning a new forecaster to a cluster in the first set based on personal characteristics of said new forecaster.
Claim 40
Dependent Claim
33 words
A method according to claim 29 , wherein said assignment formula minimizes a distance based on said personal characteristics between the new forecaster and the cluster to which the new forecaster is assigned.
Claim 41
Dependent Claim
16 words
A method according to claim 25 , wherein the forecasters comprise a large longitudinal forecast panel.
Claim 42
Dependent Claim
14 words
A method according to claim 25 , wherein the forecasters comprise thousands of people.
Claim 43
Dependent Claim
28 words
A method according to claim 25 , wherein the clusters in the first set are formed simultaneously with assigning the forecasters to said clusters in the first set.
Claim 44
Dependent Claim
71 words
A method according to claim 25 , further comprising a step of utilizing the clusters in the first set to produce combination forecasts for a variety of prediction events, wherein at least some of the forecasters participate inconsistently, and wherein for a given prediction event the statistical data for the clusters in the first set are calculated only with respect to the forecasters who have participated in said given prediction event.
Claim 45
Independent Claim
128 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and utilizing the clusters to produce combination forecasts for a variety of prediction events, wherein at least some of the forecasters participate inconsistently, and wherein for a given prediction event the statistical data for the clusters are calculated only with respect to the forecasters who have participated in said given prediction event.
Claim 46
Independent Claim
128 words
A method for providing combination forecasts, comprising: obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; dividing the forecasters into plural clusters by utilizing a statistical clustering technique; calculating statistical data for each cluster; combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and utilizing the clusters to produce combination forecasts for a variety of prediction events, wherein at least some of the forecasters participate inconsistently, and wherein for a given prediction event the statistical data for the clusters are calculated only with respect to the forecasters who have participated in said given prediction event.
Claim 47
Independent Claim
145 words
An apparatus for providing combination forecasts, said apparatus comprising: (a) means for obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; (b) means for dividing the forecasters into plural clusters by utilizing a statistical clustering technique; (c) means for calculating statistical data for each cluster; (d) means for combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and (e) means for utilizing the clusters to produce combination forecasts for a variety of prediction events, wherein at least some of the forecasters participate inconsistently, and wherein for a given prediction event the statistical data for the clusters are calculated only with respect to the forecasters who have participated in said given prediction event.
Claim 48
Independent Claim
145 words
An apparatus for providing combination forecasts, said apparatus comprising: (a) means for obtaining predictions from a group of forecasters for a value of a financial and/or economic measure that represents an aspect of an existing economic environment, wherein the forecasters comprise people; (b) means for dividing the forecasters into plural clusters by utilizing a statistical clustering technique; (c) means for calculating statistical data for each cluster; (d) means for combining the statistical data for all the clusters so as to obtain a combination forecast for the financial and/or economic measure; and (e) means for utilizing the clusters to produce combination forecasts for a variety of prediction events, wherein at least some of the forecasters participate inconsistently, and wherein for a given prediction event the statistical data for the clusters are calculated only with respect to the forecasters who have participated in said given prediction event.
References Cited (Backward Citations)
US5930762A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2007-09-14 | 2008-03-15 | 2008-09-14 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2011-09-14 | 2012-03-15 | 2012-09-14 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2015-09-14 | 2016-03-15 | 2016-09-14 | $7700.00 | Window Closed / Lapsed |