US5608620A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method of eliciting unbiased forecasts by relating a forecaster's pay to the forecaster's contribution to a collective forecast
Assignee: Individual
•
Filing: 1995-02-17
•
Granted: 1997-03-04
•
Expires: 2015-02-17
Abstract
Abstract A method of eliciting an unbiased prediction of an unknown variable value from at least one of a group of forecasters. This method of compensating individual forecasters can be applied to an entire group of forecasters so as to elicit an unbiased collective prediction. The method yields nearly unbiased predictions from risk-averse forecasters whenever at least two forecasters are employed to make the same prediction. The method involves: aggregating the predictions of the forecasters, both with and without the particular prediction of the individual forecaster; computing collective losses for both of the aggregated predictions; calculating the individual forecaster's marginal contribution to predictive accuracy, based on the difference in collective losses; and computing and paying the individual forecaster's compensation as a function of the individual's marginal contribution.
Inventors
- 👤 Carl A. Lundgren
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-05 20:25
CPC Classifications
G; G06; G06Q; G06Q10/00; G06Q10/10; G06Q10/06; G06Q10/063; G06Q10/0639; G06Q10/06398; G06Q30/00; G06Q30/02; G06Q30/0207; G06Q40/00
Claims Analysis
Total Claims: 24
Claim 1
Independent Claim
209 words
A method of soliciting unbiased forecasts of an unknown variable value from an individual forecaster and from one or more secondary forecasters, which for said individual forecaster the method comprises the steps of: a) soliciting a prediction of said unknown variable value from said individual forecaster; b) soliciting predictions of said unknown variable value from said secondary forecasters; c) aggregating the predictions of said individual forecaster and said secondary forecasters using a first aggregator function to obtain a collective prediction; d) aggregating the predictions of said secondary forecasters using a second aggregator function to obtain a secondary collective prediction; e) observing a criterion value for said unknown variable value being predicted; f) computing a collective loss using a loss function having said criterion value and said collective prediction as arguments of said loss function; g) computing a secondary collective loss using said loss function with said criterion value and said secondary collective prediction as arguments of said loss function; h) computing said individual forecaster's marginal contribution by subtracting the collective loss from the secondary collective loss; i) computing said individual forecaster's compensation amount as a monotonic transformation of said forecaster's marginal contribution; and j) paying said individual forecaster compensation equal to the value of said individual forecaster's compensation amount.
Claim 2
Independent Claim
209 words
A method of soliciting unbiased forecasts of an unknown variable value from an individual forecaster and from one or more secondary forecasters, which for said individual forecaster the method comprises the steps of: a) soliciting a prediction of said unknown variable value from said individual forecaster; b) soliciting predictions of said unknown variable value from said secondary forecasters; c) aggregating the predictions of said individual forecaster and said secondary forecasters using a first aggregator function to obtain a collective prediction; d) aggregating the predictions of said secondary forecasters using a second aggregator function to obtain a secondary collective prediction; e) observing a criterion value for said unknown variable value being predicted; f) computing a collective loss using a loss function having said criterion value and said collective prediction as arguments of said loss function; g) computing a secondary collective loss using said loss function with said criterion value and said secondary collective prediction as arguments of said loss function; h) computing said individual forecaster's marginal contribution by subtracting the collective loss from the secondary collective loss; i) computing said individual forecaster's compensation amount as a monotonic transformation of said forecaster's marginal contribution; and j) paying said individual forecaster compensation equal to the value of said individual forecaster's compensation amount.
Claim 3
Dependent Claim
14 words
A method according to claim 1 in which said compensation consists of statutory compensation.
Claim 4
Dependent Claim
14 words
A method according to claim 1 in which said compensation consists of manufactured money.
Claim 5
Dependent Claim
13 words
A method according to claim 1 in which said loss function is nonquadratic.
Claim 6
Dependent Claim
22 words
A method according to claim 1 in which said second aggregator function is a secondary aggregator function of said first aggregator function.
Claim 7
Dependent Claim
25 words
A method according to claim 1 in which said criterion value is a future observation of the actual value of said predicted unknown variable value.
Claim 8
Dependent Claim
28 words
A method according to claim 1 in which at least one of said secondary forecaster predictions was not solicited in accordance with the method of the present invention.
Claim 9
Dependent Claim
18 words
A method according to claim 1 in which said unknown variable value is the probability of an event.
Claim 10
Dependent Claim
18 words
A method according to claim 1 in which said unknown variable value is the price of a commodity.
Claim 11
Dependent Claim
16 words
A method according to claim 1 in which said unknown variable value is a corporate profit.
Claim 12
Dependent Claim
16 words
A method according to claim 1 in which said unknown variable value is a relative profit.
Claim 13
Dependent Claim
17 words
A method according to claim 1 in which said unknown variable value is a quantity of pollutant.
Claim 14
Dependent Claim
16 words
A method according to claim 1 in which said unknown variable value is a damage estimate.
Claim 15
Dependent Claim
20 words
A method according to claim 1 in which said unknown variable value is an observation of a state of nature.
Claim 16
Dependent Claim
232 words
A method according to claim 1 in which Steps a), b), c), and d) comprise the substeps of: i) soliciting a precision weight from said individual forecaster; ii) soliciting a prediction function from said individual forecaster, said prediction function having an aggregate precision weight as one of its arguments; iii) soliciting precision weights from said secondary forecasters; iv) soliciting prediction functions from said secondary forecasters, each of said prediction functions having an aggregate precision weight as one of their arguments; v) aggregating the precision weights of said individual forecaster and said secondary forecasters to obtain a collective precision weight; vi) selecting definite predictions of said individual forecaster and said secondary forecasters by using said collective precision weight as an argument of each of said prediction functions of said individual forecaster and said secondary forecasters; vii) weighting said definite predictions using said precision weights; viii) aggregating said weighted definite predictions using said first aggregator function to obtain a collective prediction; ix) aggregating the precision weights of said secondary forecasters to obtain a secondary collective precision weight; x) selecting secondary definite predictions of said secondary forecasters by using said secondary collective precision weight as an argument of each of said prediction functions of said secondary forecasters; xi) weighting said secondary definite predictions using said precision weights; xii) aggregating said weighted secondary definite predictions using said second aggregator function to obtain a secondary collective prediction.
Claim 17
Dependent Claim
19 words
A method according to claim 1 in which said criterion value is determined from at least one criterion estimate.
Claim 18
Dependent Claim
105 words
A method according to claim 11 in which the determination of said criterion estimate comprises the substeps of: i) soliciting precision weights from one or more tertiary forecasters; ii) soliciting prediction functions from said tertiary forecasters, each of said prediction functions having an aggregate precision weight as one of their arguments; iii) selecting criterion predictions of said tertiary forecasters by substituting infinity for said aggregate precision weight as said argument of each of said prediction functions of said tertiary forecasters; iv) weighting said criterion predictions using said precision weights; v) aggregating said weighted criterion predictions using a third aggregator function to obtain said criterion estimate.
Claim 19
Dependent Claim
202 words
A method according to claim 16 in which the determination of said criterion estimate comprises the substeps of: i) soliciting precision weights from one or more tertiary forecasters; ii) soliciting prediction functions from said tertiary forecasters, each of said prediction functions having an aggregate precision weight as one of their arguments; iii) selecting tertiary definite predictions of said tertiary forecasters by using a finite aggregate precision weight as said argument of each of said prediction functions of said tertiary forecasters; iv) selecting criterion predictions of said tertiary forecasters by substituting infinity for said aggregate precision weight as said argument of each of said prediction functions of said tertiary forecasters; v) ascertaining a typical functional relationship between said criterion predictions and said tertiary definite predictions; vi) if said typical functional relationship does not fall within an acceptable range that would tend to inhibit collusion, then substituting a more acceptable functional relationship to be used as said typical functional relationship; vii) computing substitute criterion predictions using said tertiary definite predictions as arguments of said typical functional relationship; viii) weighting said substitute criterion predictions using said precision weights; ix) aggregating said weighted substitute criterion predictions using a third aggregator function to obtain said criterion estimate.
Claim 20
Independent Claim
156 words
A method of soliciting unbiased forecasts of an unknown variable value from an individual forecaster and from one or more secondary forecasters, which for said individual forecaster the method comprises the steps of: a) soliciting a prediction of said unknown variable value from said individual forecaster; b) soliciting criterion predictions of said unknown variable value from said secondary forecasters; c) aggregating the criterion predictions of said secondary forecasters using an aggregator function to obtain a criterion estimate of said unknown variable value being predicted; d) computing a collective loss using a loss function having said criterion estimate and said individual prediction as arguments of said loss function; e) computing said individual forecaster's marginal contribution by subtracting said collective loss from a fixed constant; f) computing said individual forecaster's compensation amount as a monotonic transformation of said forecaster's marginal contribution; and g) paying said individual forecaster compensation equal to the value of said individual forecaster's compensation amount.
Claim 21
Independent Claim
156 words
A method of soliciting unbiased forecasts of an unknown variable value from an individual forecaster and from one or more secondary forecasters, which for said individual forecaster the method comprises the steps of: a) soliciting a prediction of said unknown variable value from said individual forecaster; b) soliciting criterion predictions of said unknown variable value from said secondary forecasters; c) aggregating the criterion predictions of said secondary forecasters using an aggregator function to obtain a criterion estimate of said unknown variable value being predicted; d) computing a collective loss using a loss function having said criterion estimate and said individual prediction as arguments of said loss function; e) computing said individual forecaster's marginal contribution by subtracting said collective loss from a fixed constant; f) computing said individual forecaster's compensation amount as a monotonic transformation of said forecaster's marginal contribution; and g) paying said individual forecaster compensation equal to the value of said individual forecaster's compensation amount.
Claim 22
Independent Claim
201 words
A method of soliciting unbiased forecasts of an unknown variable value from an individual forecaster and from one or more secondary forecasters, which for said individual forecaster the method comprises the steps of: a) announcing a preliminary prediction of said unknown variable value; b) soliciting a prediction of said unknown variable value from said individual forecaster; c) soliciting criterion predictions of said unknown variable value from said secondary forecasters; d) aggregating the criterion predictions of said secondary forecasters using an aggregator function to obtain a criterion estimate of said unknown variable value being predicted; e) computing a collective loss using a loss function having said criterion estimate and said individual prediction as arguments of said loss function; f) computing a secondary collective loss using said loss function with said criterion estimate and said preliminary prediction as arguments of said loss function; g) computing said individual forecaster's marginal contribution by subtracting said collective loss from a fixed constant and adding said secondary collective loss multiplied by a nonzero constant; h) computing said individual forecaster's compensation amount as a monotonic transformation of said forecaster's marginal contribution; and i) paying said individual forecaster compensation equal to the value of said individual forecaster's compensation amount.
Claim 23
Independent Claim
201 words
A method of soliciting unbiased forecasts of an unknown variable value from an individual forecaster and from one or more secondary forecasters, which for said individual forecaster the method comprises the steps of: a) announcing a preliminary prediction of said unknown variable value; b) soliciting a prediction of said unknown variable value from said individual forecaster; c) soliciting criterion predictions of said unknown variable value from said secondary forecasters; d) aggregating the criterion predictions of said secondary forecasters using an aggregator function to obtain a criterion estimate of said unknown variable value being predicted; e) computing a collective loss using a loss function having said criterion estimate and said individual prediction as arguments of said loss function; f) computing a secondary collective loss using said loss function with said criterion estimate and said preliminary prediction as arguments of said loss function; g) computing said individual forecaster's marginal contribution by subtracting said collective loss from a fixed constant and adding said secondary collective loss multiplied by a nonzero constant; h) computing said individual forecaster's compensation amount as a monotonic transformation of said forecaster's marginal contribution; and i) paying said individual forecaster compensation equal to the value of said individual forecaster's compensation amount.
Claim 24
Dependent Claim
13 words
A method according to claim 20 in which said nonzero constant is one.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2000-03-03 | 2000-09-02 | 2001-03-04 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2004-03-03 | 2004-09-02 | 2005-03-04 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2008-03-03 | 2008-09-02 | 2009-03-04 | $7700.00 | Window Closed / Lapsed |