CN104699595A - Software testing method facing to software upgrading - Google Patents

Software testing method facing to software upgrading Download PDF

Info

Publication number
CN104699595A
CN104699595A CN201310655851.5A CN201310655851A CN104699595A CN 104699595 A CN104699595 A CN 104699595A CN 201310655851 A CN201310655851 A CN 201310655851A CN 104699595 A CN104699595 A CN 104699595A
Authority
CN
China
Prior art keywords
transformation
software
program
particle
relation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201310655851.5A
Other languages
Chinese (zh)
Other versions
CN104699595B (en
Inventor
张洁
陈俊洁
郝丹
熊英飞
谢冰
梅宏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Peking University
Original Assignee
Peking University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Peking University filed Critical Peking University
Priority to CN201310655851.5A priority Critical patent/CN104699595B/en
Publication of CN104699595A publication Critical patent/CN104699595A/en
Application granted granted Critical
Publication of CN104699595B publication Critical patent/CN104699595B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Debugging And Monitoring (AREA)

Abstract

The invention discloses a software testing method facing to software upgrading. The method is automatically constructed on the basis of the metamorphic relation of equation description. Before software upgrading, a tester can construct the metamorphic relation of each procedure at the aim of a conventional software; the procedure refers to a procedure code for achieving a special function calculating function; after software upgrading, at the aim of a modified procedure, the tester can produce n random numbers as use cases for testing within a certain range, then substituting each case for testing and corresponding output of the case in the modified procedure into a metamorphic relation expression of the procedure before modification, and checking whether the metamorphic relation expression is tenable or not, and if the modified procedure is not conform to the metamorphic relation constructed before, errors on the software exist during the updating process. Through the method provided by the invention, the tester can apply the metamorphic relation constructed into a metamorphosis process, therefore, the correctness of the software during the version upgrading process can be verified.

Description

The method for testing software of a kind of software-oriented upgrading
Technical field
The invention provides the method for testing software automatically constructed based on the transformation relation of equation description of a kind of software-oriented upgrading, the method can assist the carrying out of metamorphic testing, belongs to software test field.
Background technology
In software test field, tester relies on Test oracles (test oracle) to judge whether program to be measured is passed through usually.But for a lot of program (as scientific program), the acquisition of Test oracles is extremely difficult, tester often cannot judge the correctness of the Output rusults under specific input.Metamorphic testing technology is a kind of novel software testing technology, and the method carrys out test procedure by the transformation relation of trace routine repeatedly between Output rusults, and the expection not needing constructor single to perform exports.
Transformation relation is the core of metamorphic testing.In brief, transformation relation table understands that the change inputted in program multiple exercise process is on the impact of Output rusults.For trigonometric function sin, sin (x+ π)=-sin (x) is that a simple transformation relation (inputs and becomes x+ π from x, then Output rusults opposite number each other), as the program P of measuring and calculation sin function, if test case is x=39 °, tester is difficult to the accurate expected results learning sin (39 °), but can be relatively easy to judge whether sin (39 ° of+π) is equal with-sin (39 °).If unequal, then there is mistake in program P.
Current metamorphic testing correlative study mainly comprises: the optimization of metamorphic testing process, the screening of transformation relation, metamorphic testing and other verify or the combination, the application of metamorphic testing technology in specific area etc. of measuring technology.Although existing research effectively can make up the deficiency of conventional test methodologies, lack practical transformation relation building method, and up to the present do not have method can realize automatically constructing transformation relation.
Summary of the invention
The present invention is directed to the deficiencies in the prior art, provide the method for testing software automatically constructed based on the transformation relation of equation description of a kind of software-oriented upgrading, can be applicable to the error detection lacking Test oracles (test oracle) program in software release upgrade process.
Technical scheme provided by the invention is as follows:
A method for testing software for software-oriented upgrading, is characterized in that,
A., before software upgrading, tester, for existing software, constructs the transformation relation of each program; Described program refers to the program code realizing specific function computing function;
B. after software upgrading, for amended program, tester produces this value of n(and can be specified by user, such as, n=100 or 200 etc. can be set) in individual certain limit the random number of (as-5 ~ 5) as test case, then in the transformation relational expression that before each test case and corresponding output in its after the modification program being substituted into amendment, program structure goes out, check whether transformation relational expression is set up, the transformation relation constructed before if amended program does not meet, then illustrate that software exists mistake in escalation process.
Wherein, in steps A, the step of constructor transformation relation is as follows:
1) transformation relation form: the transformation relation in all programs represents with following formula:
R i ( I 1 , I 2 ) ⇒ R 0 ( O 1 , O 2 ) - - - ( 1 )
Wherein, I 1for the original input of program, I 2input for a change; O 1for and I 1corresponding output, O 2for and I 2corresponding output, R irepresent I 1and I 2between relation, R 0represent O 1and O 2between relation;
2) transformation Relation Parameters: in order to construct transformation relation, by R iand R 0be limited within the scope of equation, work as R iduring for linear equality, comprise following situation:
Situation A:R ofor linear equality;
Case B: R ofor secondary equation;
Situation C:R obe three equatioies;
Suppose I 1and I 2respectively there is n group parameter value, be expressed as <x 1, x 2..., x n> and <y 1, y 2..., y n>, due to R in above three kinds of situations ibe linear equality, I 2each element can regard I as 1the linear combination of all elements, can be expressed as: wherein a ijand b ifor linear dimensions; If α representative is by a ijthe array a [i, j] of composition, β representation vector <b 1, b 2..., b n>, then I 2=α I 1+ β;
2.1) for situation A: suppose that tested program is Ρ, then O 1available P (I 1) represent, O 2available P (I 2) represent; Due to R 0also be linear equality, therefore R 0can be expressed as:
c 1P(I 1)+c 2P(aI 1+β)+d=0 (2)
Wherein, c 1, c 2, each parameter in d and a and β is transformation parameter;
2.2) for case B: R 0can be expressed as:
c 1P 2(I 1)+c 2P 2(aI 1+β)+c 3P(I 1)P(aI 1+β)+d 1P(I 1)+d 2P(aI 1+β)+e=0 (3)
Wherein, c 1, c 2, a, β, c 3, d 1, d 2, e is transformation parameter;
2.3) for situation C:R 0can be expressed as:
c 1P 3(I 1)+c 2P 3(aI 1+β)+c 3P 2(I 1)P(aI 1+β)+c 4P(I 1)P 2(aI 1+β)+c 5P 2(I 1)+c 6P 2(aI 1+β)+c 7P(I 1)P(aI 1+β)+d 1P(I 1)+d 2P(aI 1+β)+e=0 (4)
Wherein, a, β, c i(i=1..7), d j(j=1..2), e is transformation parameter;
The situation of more than three times is analogized accordingly;
3) series structure is closed in transformation: after defining transformation parameter, uses evolution algorithm (as particle cluster algorithm) to find qualified parameter combinations as much as possible.
In the present invention, the evolution algorithm of use is particle cluster algorithm (Particle Swarm Optimization).In particle cluster algorithm, each candidate solution is referred to as a particle (particle).Consider to search for N number of particle in a D dimension space, use with represent respectively i-th (1≤i≤N) individual particle time t (t=1,2 ...) speed and position.Particle is according to following formula renewal speed:
v id t + 1 = w v id t + c 1 r 1 ( p id t - l id t ) + c 2 r 2 ( p gd t - l id t ) - - - ( a )
In above formula, represent the speed in particle i t+1 moment in d dimension; W is inertia weight; c 1and c 2it is Studying factors; r 1and r 2the random number between [0,1], in order to keep the diversity of colony; it is the optimal location that particle i searches in d dimension to moment t; it is the optimal location that whole population searches up to now. be the d dimension position vector of particle i, and particle upgrades the position of oneself according to following formula:
l id t + 1 = l id t + v id t + 1 - - - ( b )
Initial time (t=1), initial velocity and the reference position of N number of particle are stochastic generation, and before arrival end time T, the speed of each particle and position upgrade according to formula (a) and formula (b).
In the present invention, by Studying factors c 1and c 2all be set to 1.49445, number of particles N location 20, total iterations (occurrence number of moment T) is decided to be 350, the inertia weight w of t t=0.9-0.5* (t/T) 2.
Evolution algorithm at every turn execution can provide one group of optimum solution, evolution algorithm is allowed to rerun repeatedly, obtain organizing optimum solution more, often organize optimum solution and be one group of transformation parameter value, transformation parameter value is brought in formula (2), (3) or (4) and transformation relational expression can be obtained.
In step 3), using clustering algorithm (as used the K-means clustering algorithm based on weighted euclidean distance), multiple similar optimum solution being merged into a solution.
Advantage of the present invention is: by the present invention, and tester can by the transformation relational application that constructs in metamorphic testing, thus the correctness of verifying software edition upgrading process Program.
Embodiment
Developer, when carrying out software release upgrade, due to operations such as interpolation, amendment, delete codes, often can make some indiscoverable mistakes.A part of code such as the implementation method of sin function is as follows:
If developer is accidentally by the parameter x in x(and Double.doubleToRawLongBits (x) of the third line) change-x into, bring to function correctness to have a strong impact on, but the method that this small mistake particular value detects cannot find (result of all particular values of sin is not by this erroneous effects); For no special value, developer and tester are difficult to again find Test oracles (test oracle).
Utilize the method for testing software that transformation relation of the present invention constructs automatically, developer only need find out the transformation relation of program needing amendment before software upgrading, verify again after program is modified new procedures whether still meet before transformation relation.
The specific embodiment of the present invention is as follows:
A) tester is for the program code (realizing the program of specific function computing function) of existing software version, uses the automatic constructing technology of transformation relation of the present invention to construct the transformation relation of each program.
B) in software upgrade process, for amended program, tester produces the random number of (as-5 ~ 5) in n certain limit as test case.
C) by each test case and its after the modification in program corresponding export substitute into amendment before in the transformation relational expression that goes out of program structure, check whether transformation relational expression is set up.The transformation relation constructed before if amended program does not meet, then illustrate that software exists mistake in escalation process.
Wherein, the detailed process constructing transformation relation is as follows:
The present invention utilizes linear relationship automatically to construct transformation relation.Concrete research step comprises transformation relation form, transformation Relation Parameters, contents such as closing series structure of changing in quality.Detailed step is as follows:
1) transformation relation form
Transformation relation table understands that the change inputted in program multiple exercise process is on the impact of Output rusults, and therefore, available formula (1) summarizes the transformation relation in all programs simply:
R i ( I 1 , I 2 ) &DoubleRightArrow; R 0 ( O 1 , O 2 ) - - - ( 1 )
In this formula, I 1for the original input of program, I 2input is for a change (by I 1conversion gets); O 1for and I 1corresponding output, O 2for and I 2corresponding output.R irepresent I 1and I 2between relation, R 0represent O 1and O 2between relation.This formula relation indicated visually between different input may cause also there is certain relation between corresponding output.
2) transformation Relation Parameters
In order to construct transformation relation, by R iand R 0be limited within the scope of equation, namely only study the building method of the transformation relation that available equation represents.The present invention now realizes R ifor linear equality, R 0for the situation of linear equality or nonlinear equation.Concrete situation is classified as follows:
Situation A:R ifor linear equality, R oalso be linear equality.
Case B: R ifor linear equality, R ofor secondary equation.
Situation C:R ifor linear equality, R obe three equatioies.
Suppose I 1and I 2respectively there is n group parameter value, be expressed as <x 1, x 2..., x n> and <y 1, y 2..., y n>.As previously noted, I 2by I 1conversion gets.Due to R in above three kinds of situations ibe linear equality, I 2each element can regard I as 1the linear combination of all elements, can be expressed as: wherein a ijand b ifor linear dimensions.If α representative is by a ijthe array a [i, j] of composition, β representation vector <b 1, b 2..., b n>, then I 2=α I 1+ β.
Consider three kinds of situations respectively below:
Situation A:
Suppose that tested program is Ρ, then O 1available P (I 1) represent, O 2available P (I 2) represent.Due to R 0also be linear equality, therefore R 0can be expressed as:
c 1P(I 1)+c 2P(aI 1+β)+d=0 (2)
Wherein, c 1, c 2, each parameter in d and a and β is transformation parameter.
Case B:
In like manner, for case B, R 0can be expressed as:
c 1P 2(I 1)+c 2P 2(aI 1+β)+c 3P(I 1)P(aI 1+β)+d 1P(I 1)+d 2P(aI 1+β)+e=0 (3)
Wherein, c 1, c 2, a, β, c 3, d 1, d 2, e is transformation parameter.
Such as, for transformation relation sin (x)+sin (x+ π)=0, from formula (2), c 1=1, c 2=1, a=1, β=π, d=0.Again such as, for transformation relation 2cos 2(x)-cos (2x)-1=0, from formula (3), c 1=2, c 2=0, c 3=0, d 1=0, d 2=-1, a=2, β=0, e=-1.
Situation C:
On this basis, for situation C, R 0can be expressed as:
c 1P 3(I 1)+c 2P 3(aI 1+β)+c 3P 2(I 1)P(aI 1+β)+c 4P(I 1)P 2(aI 1+β)+c 5P 2(I 1)+c 6P 2(aI 1+β)+c 7P(I 1)P(aI 1+β)+d 1P(I 1)+d 2P(aI 1+β)+e=0 (4)
Wherein, a, β, c i(i=1..7), d j(j=1..2), e is transformation parameter.
The situation of more than three times is feasible in theory, but transformation parameter is too many, and parametric solution exists certain difficulty, therefore the present invention does not verify the situation of more than three times.
3) series structure is closed in transformation
After defining transformation parameter, use all qualified parameter combinations of searching algorithm search.Searching algorithm at every turn execution can provide one group of optimum solution, allows searching algorithm rerun repeatedly, obtains organizing optimum solution more, often organizes solution and is one group of transformation parameter value, bring in formula (2), (3) or (4) and can obtain transformation relational expression.
For avoiding obtaining the too many optimum solution repeated, the present invention adds clustering algorithm, similar solution is merged into a solution.
It should be noted that, because the present invention exists certain false judgment (false positive), when the transformation relation of minority is breached, can not the accuracy of determination result, therefore need to set certain threshold value.Through experimental verification repeatedly, when there being the transformation relation of more than 5% to be breached, can judge to there is mistake in program.
Below by example, the present invention will be further described.
Embodiment:
The upgrading assuming that certain developer a will modify to the program realizing sin function in a mathematical function library.
Before upgrading, tester b uses the present invention to construct the transformation relation of original version program.The main transformation relation detected is as follows:
After the program realizing sin function is upgraded by developer, for detecting the rear program correctness of upgrading, whether the program after tester needs detection upgrading meets each the transformation relation in upper showing.Such as transformation relation sin (x)-sin (x-2 π)=0, tester needs random number between generation 100-5 ~ 5 as test case, these 100 test cases are brought in transformation relational expression sin (x)-sin (x-2 π)=0 and checks whether equation is still set up, if certain test case makes equation be false, this test case is described not by test.Other transformation relational expressions are processed too.
Owing to there is certain false judgment (false positive), when the test case of minority is by test, can not the accuracy of determination result, therefore need to judge whether there is mistake in program according to threshold value 5%.Such as, in this example, for some specific transformation relations, more than 5 in 100 random test use-cases, are had obstructedly out-of-dately just can to judge that this transformation relation is breached.As long as have one to be breached in all transformation relations can judge to there is mistake in program.

Claims (7)

1. a method for testing software for software-oriented upgrading, is characterized in that,
A., before software upgrading, tester, for existing software, constructs the transformation relation of each program; Described program refers to the program code realizing specific function computing function;
B. after software upgrading, for amended program, tester produces random number in n certain limit as test case, then in the transformation relational expression that before each test case and corresponding output in its after the modification program being substituted into amendment, program structure goes out, check whether transformation relational expression is set up, the transformation relation constructed before if amended program does not meet, then illustrate that software exists mistake in escalation process.
2. method for testing software as claimed in claim 1, it is characterized in that, in steps A, the step of constructor transformation relation is as follows:
1) transformation relation form: the transformation relation in all programs represents with following formula:
R i ( I 1 , I 2 ) &DoubleRightArrow; R 0 ( O 1 , O 2 ) - - - ( 1 )
Wherein, I 1for the original input of program, I 2input for a change; O 1for and I 1corresponding output, O 2for and I 2corresponding output, R irepresent I 1and I 2between relation, R 0represent O 1and O 2between relation;
2) transformation Relation Parameters: in order to construct transformation relation, by R iand R 0be limited within the scope of equation, work as R iduring for linear equality, comprise following situation:
Situation A:R ofor linear equality;
Case B: R ofor secondary equation;
Situation C:R obe three equatioies;
Suppose I 1and I 2respectively there is n group parameter value, be expressed as <x 1, x 2..., x n> and <y 1, y 2..., y n>, due to R in above three kinds of situations ibe linear equality, I 2each element can regard I as 1the linear combination of all elements, can be expressed as: wherein a ijand b ifor linear dimensions; If α representative is by a ijthe array a [i, j] of composition, β representation vector <b 1, b 2..., b n>, then I 2=α I 1+ β;
2.1) for situation A: suppose that tested program is Ρ, then O 1available P (I 1) represent, O 2available P (I 2) represent; Due to R 0also be linear equality, therefore R 0can be expressed as:
c 1P(I 1)+c 2P(aI 1+β)+d=0 (2)
Wherein, c 1, c 2, each parameter in d and a and β is transformation parameter;
2.2) for case B: R 0can be expressed as:
c 1P 2(I 1)+c 2P 2(aI 1+β)+c 3P(I 1)P(aI 1+β)+d 1P(I 1)+d 2P(aI 1+β)+e=0 (3)
Wherein, c 1, c 2, a, β, c 3, d 1, d 2, e is transformation parameter;
2.3) for situation C:R 0can be expressed as:
c 1P 3(I 1)+c 2P 3(aI 1+β)+c 3P 2(I 1)P(aI 1+β)+c 4P(I 1)P 2(aI 1+β)+c 5P 2(I 1)+c 6P 2(aI 1+β)+c 7P(I 1)P(aI 1+β)+d 1P(I 1)+d 2P(aI 1+β)+e=0 (4)
Wherein, a, β, c i(i=1..7), d j(j=1..2), e is transformation parameter;
The situation of more than three times is analogized accordingly;
3) series structure is closed in transformation: after defining transformation parameter, evolution algorithm is used to find qualified parameter combinations as much as possible, evolution algorithm at every turn execution can provide one group of optimum solution, evolution algorithm is allowed to rerun repeatedly, obtain organizing optimum solution more, often organize optimum solution and be one group of transformation parameter value, transformation parameter value is brought in formula (2), (3) or (4) and transformation relational expression can be obtained.
3. method for testing software as claimed in claim 2, is characterized in that, in step 3), uses clustering algorithm, multiple similar optimum solution is merged into a solution.
4. method for testing software as claimed in claim 2, is characterized in that, when the transformation relation exceeding appointment threshold value is breached, can judge to there is mistake in program.
5. method for testing software as claimed in claim 4, it is characterized in that, described threshold value is 5%.
6. method for testing software as claimed in claim 2, it is characterized in that, the evolution algorithm used in step 3) is particle cluster algorithm: in particle cluster algorithm, and each candidate solution is referred to as a particle, considers to search for N number of particle in a D dimension space, uses with represent respectively i-th (1≤i≤N) individual particle time t (t=1,2 ...) speed and position; Particle is according to following formula renewal speed:
v id t + 1 = w v id t + c 1 r 1 ( p id t - l id t ) + c 2 r 2 ( p gd t - l id t ) - - - ( a )
In above formula, represent the speed in particle i t+1 moment in d dimension; W is inertia weight; c 1and c 2it is Studying factors; r 1and r 2the random number between [0,1], in order to keep the diversity of colony; it is the optimal location that particle i searches in d dimension to moment t; it is the optimal location that whole population searches up to now; be the d dimension position vector of particle i, and particle upgrades the position of oneself according to following formula:
l id t + 1 = l id t + v id t + 1 - - - ( b )
Initial time, t=1, initial velocity and the reference position of N number of particle are stochastic generation, and before arrival end time T, the speed of each particle and position upgrade according to formula (a) and formula (b).
7. method for testing software as claimed in claim 6, is characterized in that, by Studying factors c 1and c 2all be set to 1.49445, number of particles N location 20, total iterations is decided to be 350, the inertia weight w of t t=0.9-0.5* (t/T) 2.
CN201310655851.5A 2013-12-06 2013-12-06 A kind of method for testing software of software-oriented upgrading Active CN104699595B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310655851.5A CN104699595B (en) 2013-12-06 2013-12-06 A kind of method for testing software of software-oriented upgrading

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310655851.5A CN104699595B (en) 2013-12-06 2013-12-06 A kind of method for testing software of software-oriented upgrading

Publications (2)

Publication Number Publication Date
CN104699595A true CN104699595A (en) 2015-06-10
CN104699595B CN104699595B (en) 2017-07-28

Family

ID=53346751

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310655851.5A Active CN104699595B (en) 2013-12-06 2013-12-06 A kind of method for testing software of software-oriented upgrading

Country Status (1)

Country Link
CN (1) CN104699595B (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105512046A (en) * 2016-02-01 2016-04-20 北京理工大学 Particle swarm optimization (PSO) algorithm based Android automatic testing method
CN107562265A (en) * 2017-08-22 2018-01-09 惠州Tcl移动通信有限公司 Mobile terminal and touch-screen Debugging message control process method and storage medium
CN109766259A (en) * 2018-12-03 2019-05-17 北京信息科技大学 A kind of classifier test method and system based on compound transformation relationship
CN111008139A (en) * 2019-12-06 2020-04-14 北京京航计算通讯研究所 Auxiliary method for analyzing influence domain of software change
CN111061634A (en) * 2019-12-06 2020-04-24 北京京航计算通讯研究所 Software change impact domain analysis auxiliary system
CN112817846A (en) * 2021-01-27 2021-05-18 北京科技大学 Metamorphic testing method for concurrent programs

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102053910A (en) * 2010-11-18 2011-05-11 西北工业大学 Embedded software testing method based on AADL (Architecture Analysis and Design Language) mode transformation relationship

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102053910A (en) * 2010-11-18 2011-05-11 西北工业大学 Embedded software testing method based on AADL (Architecture Analysis and Design Language) mode transformation relationship

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
T.Y.CHEN 等: "Fault-Based Testing Without the Need of Oracles", 《INFORMATION AND SOFTWARE TECHNOLOGY》 *
董国伟 等: "蜕变测试技术综述", 《计算机科学与探索》 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105512046A (en) * 2016-02-01 2016-04-20 北京理工大学 Particle swarm optimization (PSO) algorithm based Android automatic testing method
CN107562265A (en) * 2017-08-22 2018-01-09 惠州Tcl移动通信有限公司 Mobile terminal and touch-screen Debugging message control process method and storage medium
CN109766259A (en) * 2018-12-03 2019-05-17 北京信息科技大学 A kind of classifier test method and system based on compound transformation relationship
CN109766259B (en) * 2018-12-03 2022-02-01 北京信息科技大学 Classifier testing method and system based on composite metamorphic relation
CN111008139A (en) * 2019-12-06 2020-04-14 北京京航计算通讯研究所 Auxiliary method for analyzing influence domain of software change
CN111061634A (en) * 2019-12-06 2020-04-24 北京京航计算通讯研究所 Software change impact domain analysis auxiliary system
CN112817846A (en) * 2021-01-27 2021-05-18 北京科技大学 Metamorphic testing method for concurrent programs
CN112817846B (en) * 2021-01-27 2023-08-08 北京科技大学 Metamorphic testing method for concurrent program

Also Published As

Publication number Publication date
CN104699595B (en) 2017-07-28

Similar Documents

Publication Publication Date Title
CN104699595A (en) Software testing method facing to software upgrading
Gamboa et al. Statistical inference for Sobol pick-freeze Monte Carlo method
CN111900731A (en) PMU-based power system state estimation performance evaluation method
Wu et al. Review on evaluation criteria of machine learning based on big data
CN105808962A (en) Assessment method considering voltage probabilities of multiple electric power systems with wind power output randomness
CN117035563B (en) Product quality safety risk monitoring method, device, monitoring system and medium
CN103207804A (en) MapReduce load simulation method based on cluster job logging
Caprini et al. Functional-analysis based tool for testing quark-hadron duality
Wu et al. Optimization and improvement based on K-Means Cluster algorithm
Zhao et al. Threshold dynamics of the stochastic epidemic model with jump-diffusion infection force
CN104899371B (en) Survey of Transmission Line method and device
Zhang et al. Modelling Quality Attributes in Feature Models in Software Product Line Engineering.
Chen et al. Research on Audit Simulation of Accounting Computerization Based on Internet Complex Discrete Dynamic Modeling Technology
CN113919609A (en) Power distribution network model quality comprehensive evaluation method and system
Klyatis et al. Multi-variate Weibull model for predicting system-reliability, from testing results of the components
CN106407274A (en) Model assessment method and apparatus
CN113688178B (en) Footprint accounting method and system
CN102426631A (en) High-dimension space mapping-based K harmonic mean clustering method
Shen et al. Research on the Application of Data Mining in Software Testing and Defects Analysis
Zhang et al. The improvement of Gk-tail algorithm of software behavior modeling
Qin et al. Construction and application of financial fraud early warning model based on data mining
Jiangang et al. Research of Software Testing Model based on Correlation Defect
Zhou et al. Program completeness verification mechanism based on static analysis
Li et al. Automatic classification of multi-source and multi-granularity teaching resources based on random forest algorithm
Yu et al. A Model Library Method of Android Malware Detection for Population

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant