US 11,868,425 B2
Dynamic outlier bias reduction system and method
Richard B. Jones, Georgetown, TX (US)
Assigned to HARTFORD STEAM BOILER INSPECTION AND INSURANCE COMPANY, Hartford, CT (US)
Filed by HARTFORD STEAM BOILER INSPECTION AND INSURANCE COMPANY, Hartford, CT (US)
Filed on May 16, 2022, as Appl. No. 17/745,596.
Application 17/745,596 is a continuation of application No. 15/963,817, filed on Apr. 26, 2018, granted, now 11,334,645.
Application 15/963,817 is a continuation of application No. 14/738,266, filed on Jun. 12, 2015, abandoned.
Application 14/738,266 is a continuation of application No. 13/213,780, filed on Aug. 19, 2011, granted, now 9,069,725, issued on Jun. 30, 2015.
Prior Publication US 2022/0277058 A1, Sep. 1, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 17/10 (2006.01); G06F 17/18 (2006.01); G01N 33/00 (2006.01); G06N 5/04 (2023.01); G06F 7/58 (2006.01)
CPC G06F 17/10 (2013.01) [G01N 33/0004 (2013.01); G06F 17/18 (2013.01); G06N 5/04 (2013.01); G06F 7/588 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
electronically receiving, by a processor, at least the following:
a model for one or more operating conditions, and
facility operating data of the one or more operating conditions for each respective facility of a plurality of facilities;
wherein the model comprises one or more coefficients;
determining, by the processor, one or more acceptable bias criteria associated with the model;
iteratively performing, by the processor, one or more iterations of outlier bias reduction in the facility operating data of the plurality of facilities based at least in part on the model;
wherein the iteratively performing the one or more iterations of outlier bias reduction comprises steps of:
(i) determining a set of model-predicted values;
(ii) comparing the set of model-predicted values to the facility operating data to produce a set of error values;
(iii) removing bias facility operating data of one or more performance outlier facilities from the facility operating data of the plurality of facilities to form a non-biased facility operating data of one or more performance non-biased facilities that are selected from the plurality of facilities, wherein the one or more performance outlier facilities are determined based at least in part on the set of error values and the one or more acceptable bias criteria;
(iv) constructing, based at least in part on the non-biased facility operating a data, an updated model for the one or more operating conditions, wherein the updated model comprises one or more updated coefficients;
(v) repeating steps (i) through (iv) when one or more termination criteria are not satisfied;
determining, by the processor, based at least in part on the non-biased facility operating data for the one or more operating conditions of the one or more performance non-biased facilities, one or more non-biased performance standards for the one or more operating conditions; and
tracking, by processor, based at least in part on the one or more non-biased performance standards and the facility operating data, operating performance of each respective facility of the plurality of facilities so that operating performance comports with the one or more non-biased performance standards.