Data analysis apparatus, data analysis method, and computer products

• US 20020170022A1
• Filed: 01/30/2002
• Published: 11/14/2002
• Est. Priority Date: 04/25/2001
• Status: Active Grant

• Current Assignee: Socionext Inc.
• Original Assignee: Fujitsu Limited
• Inventors: Tsuda, Hidetaka, Shirai, Hidehiro
• Granted Patent: US 6,711,522 B2
• Time in Patent Office: Days
• US Class Current: 716/4
• 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: H01L 22/20 Sequence of activities consisting of a plurality of measurements, corrections, marking or sorting steps

Abstract

Data to be analyzed, for example, a yield value and various measurement values are selected and extracted from an original data group (step S1). At least one data distribution characteristic is extracted with respect to the extracted data (step S2). The data distribution characteristic amount to be analyzed is selected from these characteristics, and data mining such as a regression tree analysis is performed, designating it as a target variable (step S3). After the regression tree analysis has been completed with respect to all the extracted data distribution characteristics (step S4) the analysis result is output, and the engineer confirms it (step S5), and makes a decision (step S6).

Claims

1. A data analysis method comprising the steps of:

quantitatively evaluating and extracting at least one data distribution characteristic amount existing in an original data group, by editing the original data value;

selecting an optional data distribution characteristic amount from the extracted data distribution characteristic amount and performing an analysis; and

making a decision based on the obtained analysis result.

2. The data analysis method according to claim 1, wherein the data distribution characteristic amount is shown in a continuous value representing the degree of characteristic.

3. The data analysis method according to claim 1, wherein each data distribution characteristic amount is independent of each other, for each record.

4. The data analysis method according to claim 1, wherein the analysis is performed by data mining in which the data distribution characteristic amount is designated as a target variable.

5. The data analysis method according to claim 4, wherein individual data distribution characteristic amount is stored in a file for each record, and the same data distribution characteristic amount is sequentially selected from the file, with respect to a part of or all of the records, and designated as the target variable, to perform a regression tree analysis.

6. The data analysis method according to claim 1, wherein the each step is automatically performed by a computer system which executes software incorporated therein so as to be performed sequentially.

7. The data analysis method according to claim 1, wherein one of the data distribution characteristic amount is a value “

a”

of a y-axis section at the time of approximating the relation between x and y to a linear equation y=b·

x+a, designating an order of the original data array as x, and an original data value as y.

8. The data analysis method according to claim 1, wherein one of the data distribution characteristic amount is a value “

b”

of an inclination at the time of approximating the relation between x and y to a linear equation y=b·

x+a, designating an order of the original data array as x, and an original data value as y.

9. The data analysis method according to claim 1, wherein one of the data distribution characteristic amount is a specific periodicity strength of the original data value with respect to the order of the original data array.

10. The data analysis method according to claim 1, wherein one of the data distribution characteristic amount is a value indicating the strongest periodicity of the original data value with respect to the order of the original data array.

11. A data analysis method comprising the steps of:

quantitatively evaluating and extracting at least two data distribution characteristic amounts existing in an original data group, by editing the original data value;

sequentially selecting the extracted individual data distribution characteristic amount and performing an analysis; and

making a decision based on the obtained analysis result.

12. The data analysis method according to claim 11, wherein the data distribution characteristic amount is shown in a continuous value representing the degree of characteristic.

13. The data analysis method according to claim 11, wherein each data distribution characteristic amount is independent of each other, for each record.

14. The data analysis method according to claim 11, wherein the analysis is performed by data mining in which the data distribution characteristic amount is designated as a target variable.

15. The data analysis method according to claim 14, wherein individual data distribution characteristic amount is stored in a file for each record, and the same data distribution characteristic amount is sequentially selected from the file, with respect to a part of or all of the records, and designated as the target variable, to perform a regression tree analysis.

16. The data analysis method according claim 11, wherein the each step is automatically performed by a computer system which executes software incorporated therein so as to be performed sequentially.

17. The data analysis method according to claim 11, wherein one of the data distribution characteristic amount is a value “

a”

of ay-axis section at the time of approximating the relation between x and y to a linear equation y=b·

x+a, designating an order of the original data array as x, and an original data value as y.

18. The data analysis method according to claim 11, wherein one of the data distribution characteristic amount is a value “

b”

of an inclination at the time of approximating the relation between x and y to a linear equation y=b·

x+a, designating an order of the original data array as x, and an original data value as y.

19. The data analysis method according to claim 11, wherein one of the data distribution characteristic amount is a specific periodicity strength of the original data value with respect to the order of the original data array.

20. The data analysis method according to claim 11, wherein one of the data distribution characteristic amount is a value indicating the strongest periodicity of the original data value with respect to the order of the original data array.

21. A data analysis method comprising:

(a) a step of preparing a data result of an explanatory variable and a target variable;

(b) a step of calculating the confounding degree and/or the independence degree between a plurality of explanatory variables based on the data result; and

(c) a step of performing data mining, using the confounding degree and/or the independence degree.

22. The data analysis method according to claim 21, wherein the step (b) calculates the confounding degree and/or the independence degree in a unit of set bisected by a regression tree analysis.

23. The data analysis method according to claim 22, wherein the step (b) selects a plurality of explanatory variables which becomes a factor causing a bisection having a large significant difference by the regression tree analysis, and calculates the confounding degree and/or the independence degree between the plurality of explanatory variables.

24. The data analysis method according to claim 23, wherein the step (b) calculates the confounding degree and/or the independence degree based on a ratio of agreement of data to disagreement of data between explanatory variables in each set bisected by the regression tree analysis, at the time of calculating the confounding degree and/or the independence degree between a reference explanatory variable and other explanatory variables.

25. The data analysis method according to claim 24, wherein the step (c) performs data mining by picking up an explanatory variables based on the confounding degree and/or the independence degree.

27. The data analysis apparatus according to claim 26, wherein the calculation unit calculates the confounding degree and/or the independence degree in a unit of set bisected by means of a regression tree analysis.

28. The data analysis apparatus according to claim 27, wherein the calculation unit selects a plurality of explanatory variables which becomes a factor causing a bisection having a large significant difference by the regression tree analysis, and calculates the confounding degree and/or the independence degree between the plurality of explanatory variables.

29. The data analysis apparatus according to claim 27, wherein the calculation unit calculates the confounding degree and/or the independence degree based on a ratio of agreement of data to disagreement of data between explanatory variables in each set bisected by the regression tree analysis, at the time of calculating the confounding degree and/or the independence degree between a reference explanatory variable and other explanatory variables.

30. The data analysis apparatus according to claim 29, wherein the data mining unit performs data mining by picking up an explanatory variables based on the confounding degree and/or the independence degree.

26. A data analysis apparatus comprising:

a calculation unit which calculates a confounding degree and/or an independence degree between a plurality of explanatory variables, based on the data result of the explanatory variables and a target variable; and

a data mining unit which performs data mining by using the confounding degree and/or the independence degree.

31. A computer readable recording medium which records a program for a computer to execute, the program comprising:

(a) a procedure of preparing a data result of an explanatory variable and a target variable;

(b) a procedure of calculating the confounding degree and/or the independence degree between a plurality of explanatory variables based on the data result; and

(c) a procedure of performing data mining, using the confounding degree and/or the independence degree.

32. A computer program for a computer to execute, the program comprising:

(a) a procedure of preparing a data result of an explanatory variable and a target variable;

(b) a procedure of calculating the confounding degree and/or the independence degree between a plurality of explanatory variables based on the data result; and

(c) a procedure of performing data mining, using the confounding degree and/or the independence degree.