Statistical pattern recognition and analysis

• US 20070136115A1
• Filed: 12/13/2005
• Published: 06/14/2007
• Est. Priority Date: 12/13/2005
• Status: Abandoned Application

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: LaComb, Christina, Vivier, Barbara, Senturk Doganaksoy, Deniz
• Application Number:
  US11/301,669
• Publication Number:
  US 20070136115A1
• Time in Patent Office: Days
• US Class Current: 705/7
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes

Abstract

A technique is provided for analyzing a dataset. The technique includes generating multivariate parameters to capture statistical patterns over time and/or across dimensions in the dataset, and developing a dynamic model based on the multivariate parameters for analyzing the dataset.

Claims

1. A method of capturing statistical patterns in a dataset, the method comprising:

(a) representing time-varying and/or dimension-varying data in the dataset using statistics; and

(b) deriving multivariate parameters based on the statistical data, the multivariate parameters being indicative of statistical patterns in the dataset.

2. The method of claim 1, wherein step (a) comprises representing time-varying data using moving averages, moving medians, moving quartiles, or moving standard deviations.

3. The method of claim 1, wherein step (a) comprises aggregating the dimension-varying data via central tendency, variance or Z-score.

4. The method of claim 1, wherein dimension-varying data comprises two or more metrics, red flags, rule based categorical measures, and/or discrete quantities.

5. The method of claim 1, wherein step (b) comprises computing ratio of total number of metrics that exceed the negative threshold of the modified Z-scores across the given period of time and given set of metrics to the number of non-missing Z-scores across the given period of time and given set of metrics.

6. The method of claim 1, wherein the dataset comprises a financial dataset, a demographic dataset, a behavioral dataset or a census dataset.

7. A method of analyzing a dataset, the method comprising:

generating multivariate parameters to capture statistical patterns over time and/or across dimensions in the dataset; and

developing a dynamic model based on the multivariate parameters for analyzing the dataset.

8. The method of claim 7, wherein the statistical patterns represent analytical measures and/or rule-based measures.

9. The method of claim 7, wherein the dataset comprises quantitative and/or qualitative dataset.

10. The method of claim 7, further comprising analyzing the dataset to detect an anomaly in the dataset via an anomaly detection technique.

11. The system of claim 10, wherein the anomaly detection technique comprises at least one of outlier detection, trend analysis, correlation analysis, regression analysis, and factor and cluster analysis.

12. The method of claim 10, further comprising generating an alert signal, wherein the alert signal comprises at least one of a visual representation and textual representation of the detected anomaly.

13. A method of assessing financial health of a business entity, the method comprising:

acquiring patterns statistically over time and/or across dimensions, the patterns representing financial data and/or business data related to the business entity;

developing a dynamic model based on the acquired patterns for analyzing financial and/or business data; and

assessing financial health of the business entity based on the dynamic model.

14. The method of claim 13, further comprises analyzing the financial data and/or business data using the financial anomaly detection technique to detect the behavioral patterns associated with the business entity.

15. The method of claim 14, wherein the behavioral patterns comprise at least one of likelihood of fraud, financial credit or investment risk and good credit or investment prospect associated with the business entity.

16. A system for capturing statistical patterns in a dataset, the system comprising:

a processor configured to represent time-varying and/or dimension-varying data in the dataset using statistics, and to derive multivariate parameters based on the statistical data, the multivariate parameters being indicative of statistical patterns in the dataset.

17. A data analysis system, comprising:

a processor configured to generate multivariate parameters to capture statistical patterns over time and/or across dimensions in the dataset, and to develop a dynamic model based on the multivariate parameters for analyzing the dataset.

18. The data analysis system of claim 17, wherein the processor is further configured to analyze the dataset to detect an anomaly in the dataset via an anomaly detection technique.

19. A computer readable media, comprising:

routines for representing time-varying and/or dimension-varying data in the dataset using statistics; and

routines for deriving multivariate parameters based on the statistical data, the multivariate parameters being indicative of statistical patterns in the dataset.

20. A computer readable media, comprising:

routines for generating multivariate parameters to capture statistical patterns over time and/or across dimensions in the dataset; and

routines for developing a dynamic model based on the multivariate parameters for analyzing the dataset.