Method, device and medium for checking consistency of multi-source probability test data based on binomial distribution

Abstract

The disclosed embodiments disclose a method, device and medium for checking the consistency of multi-source probability test data based on binomial distribution. The method may include: for two groups of sample data that obey the binomial distribution, respectively calculating the corresponding success probabilities according to the sample size and the number of successes; setting thresholds for the test deviation, the risk of rejecting a true false positive and the risk of retaining a false positive for the two groups of sample data, and obtaining a risk ratio according to the thresholds for the risk of rejecting a true false positive and the risk of retaining a false positive; determining a test parameter value according to the comparison results of the sample size in the two groups of sample data and the success probabilities; calculating a test threshold according to the risk ratio and the test parameter value; and comparing the test threshold with the test parameter value to determine the consistency of the two groups of sample data.

Description

Technical Field

[0001] The disclosed embodiments relate to the technical field of consistency verification of probability indicators, and in particular to a method, device and medium for consistency verification of multi-source probability test data based on binomial distribution. Background Art

[0002] Multi-source probability test data refers to probability data from multiple different sources or experimental environments. These data are usually used to evaluate the consistency test or reliability between different data sources. The consistency test refers to the test used in statistics to test whether two groups of sample data or multiple groups of sample data come from the same population or have similar characteristics. For probability indicators based on binomial distribution, it means that when a series of independent repeated binary experiments are conducted, the probability distribution of the number of successes of the probability indicator is described and calculated by binomial distribution.

[0003] In order to evaluate the consistency of the internal and external data of the probability indicators of the missile based on binomial distribution, the traditional consistency verification methods usually rely on empirical rules or simple statistical techniques, but these methods often cannot provide sufficient scientific basis to evaluate the consistency of the internal and external data. For the consistency verification evaluation of probability indicators based on binomial distribution, there is a lack of scientific consistency verification methods for internal and external data, and it is impossible to associate the consistency verification with verification bias, true rejection risk and false existence risk, which leads to the inability to effectively associate and utilize the internal and external data, thereby affecting the accuracy and credibility of the evaluation results. Summary of the invention

[0004] In view of this, the embodiments of the present disclosure hope to provide a method, device and medium for consistency verification of multi-source probability test data based on binomial distribution, which can accurately verify the consistency of indoor data and outdoor data, and associate the consistency check with the verification deviation, the risk of rejecting true and the risk of existing false, so as to ensure the accuracy and reliability of the consistency verification results.

[0005] The technical solution of the embodiment of the present disclosure is implemented as follows:

[0006] In a first aspect, an embodiment of the present disclosure provides a method for checking consistency of multi-source probability test data based on binomial distribution, comprising:

[0007] For two groups of sample data that follow binomial distribution, the corresponding success probabilities are calculated based on the sample size and the number of successes.

[0008] Setting thresholds of test deviation, risk of rejecting a true result, and risk of retaining a false result for the two groups of sample data, and obtaining a risk ratio according to the thresholds of risk of rejecting a true result and risk of retaining a false result;

[0009] Determining a test parameter value according to a comparison result of the sample size in the two groups of sample data and the success probability;

[0010] Calculate and obtain a test threshold value according to the risk ratio and the test parameter value;

[0011] The inspection threshold is compared with the inspection parameter value to determine the consistency of the two sets of sample data.

[0012] In a second aspect, the embodiment of the present disclosure provides a device for checking consistency of multi-source probability test data based on binomial distribution, the device comprising: a calculation part, a setting part, a determination part, an acquisition part and a judgment part; wherein,

[0013] The calculation part is configured to calculate the corresponding success probabilities for two groups of sample data that obey the binomial distribution according to the sample size and the number of successes;

[0014] The setting part is configured to set the thresholds of the inspection deviation, the risk of rejecting a true false positive and the risk of excluding a false positive of the two groups of sample data, and obtain the risk ratio according to the thresholds of the risk of rejecting a true false positive and the risk of excluding a false positive;

[0015] The determining part is configured to determine the inspection parameter value according to the sample size in the two sets of sample data and the comparison result of the success probability;

[0016] The obtaining part is configured to calculate and obtain a test threshold value according to the risk ratio and the test parameter value;

[0017] The determination section is configured to compare the inspection threshold with the inspection parameter value to determine the consistency of the two sets of sample data.

[0018] In a third aspect, an embodiment of the present disclosure provides a computing device, the computing device comprising: a memory and a processor; wherein:

[0019] The memory is used to store a computer program that can be run on the processor;

[0020] The processor is used to execute the multi-source probability test data consistency verification method based on binomial distribution described in the first aspect when running the computer program.

[0021] In a fourth aspect, an embodiment of the present disclosure provides a computer storage medium storing at least one instruction, wherein the at least one instruction is used to be executed by a processor to implement the multi-source probability test data consistency verification method based on binomial distribution described in the first aspect.

[0022] The disclosed embodiment provides a method, device and medium for checking consistency of multi-source probability test data based on binomial distribution, by calculating the corresponding success probability according to the sample size and the number of successes for two groups of sample data that obey the binomial distribution, wherein the first group of sample data represents the digital simulation experiment results, i.e., the infield data, and the second group of sample data represents the semi-physical simulation test results, i.e., the outfield data. The thresholds of the inspection deviation, the risk of rejecting the truth and the risk of false existence of the two groups of sample data are pre-set, and the risk ratio is obtained according to the thresholds of the risk of rejecting the truth and the risk of false existence, and the corresponding success probability is calculated for the two groups of sample data, and the inspection parameter value is determined according to the comparison result of the sample size and the success probability in the two groups of sample data, and the inspection threshold is calculated according to the risk ratio and the inspection parameter value, and the inspection threshold is compared with the inspection parameter value to judge the consistency of the two groups of sample data. Through this technical solution, binomial distribution is used as a probability model, and consistency verification is associated with inspection deviation, risk of rejecting the truth and risk of false existence, so that the consistency of the infield data and the outfield data can be accurately verified, and the accuracy and reliability of the consistency verification result are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A flow chart of a method for checking consistency of multi-source probability test data based on binomial distribution provided in an embodiment of the present disclosure;

[0024] Figure 2 A detailed flow chart of a method for checking consistency of multi-source probability test data based on binomial distribution provided in an embodiment of the present disclosure;

[0025] Figure 3 A schematic diagram of a multi-source probability test data consistency verification device based on binomial distribution provided by an embodiment of the present disclosure;

[0026] Figure 4 A schematic diagram of the hardware structure of a computing device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure.

[0028] Generally speaking, due to the differences in test conditions, test equipment, the influence of environmental factors, and the complexity and variability of the field environment, it becomes difficult to simulate and reproduce the actual operating conditions, which increases the difficulty of evaluating inspection deviations and risks, and also leads to the high cost and difficulty of obtaining infield data and field data, and it is impossible to obtain enough field data for comprehensive statistical analysis, wherein the infield data is the data obtained in the laboratory environment, and the field data is the data obtained in the actual real environment. In addition, the theoretical model of the field cannot fully capture all the factors affecting the actual operation. When considering the inspection deviation, the risk of discarding the true and the risk of false existence, the simplification of the model will lead to deviations between the evaluation results and the actual situation. Therefore, the consistency verification evaluation of probability indicators based on binomial distribution currently faces a common problem, which is the lack of scientific consistency verification methods for infield data and field data, and the inability to associate consistency verification with inspection deviation, the risk of discarding the true and the risk of false existence, resulting in the inability to effectively associate and utilize the infield data and field data, thereby affecting the accuracy and credibility of the evaluation results. Based on this, the embodiment of the present disclosure hopes to provide a technical solution for consistency check of multi-source probability test data based on binomial distribution, which integrates the principles of probability theory and statistics, can accurately verify the consistency of indoor data and outdoor data, and associate the consistency check with factors such as test deviation, risk of rejection and risk of false existence, so as to ensure the accuracy and reliability of the consistency check results. Figure 1 , which shows a method for checking consistency of multi-source probability test data based on binomial distribution provided by an embodiment of the present disclosure, the method comprising:

[0029] S101: For two groups of sample data that obey binomial distribution, the corresponding success probabilities are calculated according to the sample size and the number of successes;

[0030] S102: setting thresholds of the inspection deviation, the risk of rejecting a true false positive, and the risk of retaining a false positive for the two groups of sample data, and obtaining a risk ratio according to the thresholds of the risk of rejecting a true false positive and the risk of retaining a false positive;

[0031] S103: Determine a test parameter value according to the sample size in the two groups of sample data and the comparison result of the success probability;

[0032] S104: Calculating a test threshold according to the risk ratio and the test parameter value;

[0033] S105: Compare the inspection threshold with the inspection parameter value to determine the consistency of the two groups of sample data.

[0034] According to the description of the above scheme, the embodiment of the present disclosure calculates the corresponding success probability for two groups of sample data that obey the binomial distribution according to the sample size and the number of successes, wherein the first group of sample data represents the results of the digital simulation experiment, i.e., the infield data, and the second group of sample data represents the results of the semi-physical simulation test, i.e., the outfield data. The thresholds of the inspection deviation, the risk of rejecting the truth, and the risk of false existence of the two groups of sample data are pre-set, and the risk ratio is obtained according to the thresholds of the risk of rejecting the truth and the risk of false existence, and the corresponding success probability is calculated for the two groups of sample data, and the inspection parameter value is determined according to the sample size in the two groups of sample data and the comparison result of the success probability, and the inspection threshold is calculated according to the risk ratio and the inspection parameter value, and the inspection threshold is compared with the inspection parameter value to judge the consistency of the two groups of sample data. Through this technical solution, using the binomial distribution as a probability model, the consistency check is associated with the inspection deviation, the risk of rejecting the truth, and the risk of false existence, so that the consistency of the infield data and the outfield data can be accurately verified, and the accuracy and reliability of the consistency check result are improved.

[0035] against Figure 1 In some possible implementations of the technical solution shown, the two groups of sample data that obey the binomial distribution are respectively calculated to obtain corresponding success probabilities according to the sample size and the number of successes, including:

[0036] Divide the number of successes of the first set of sample data by the sample size to obtain the corresponding probability of success of the first set of sample data;

[0037] The number of successes of the second set of sample data is divided by the sample size to obtain the corresponding probability of success of the second set of sample data.

[0038] For the above implementation, specifically, two groups of sample data of probability indicators that obey the binomial distribution are obtained, and each group of sample data includes the sample size and the number of successes, wherein the binomial distribution is a discrete probability distribution, which is used to describe the probability distribution of the number of successes in a series of independent repeated experiments, each experiment is independent of each other, and the test results have only two possibilities of success or failure, and the probability of success is fixed. Of the two groups of sample data, the first group of sample data is used to characterize the results of digital simulation experiments, i.e., indoor data, and the second group of sample data is used to characterize the results of semi-physical simulation experiments, i.e., outdoor data. The first group of sample data obeys the binomial distribution , the second set of sample data follows the binomial distribution ,in, is a random variable, is the number of trials, , is the success probability. The first set of sample data obtained is , the second set of sample data is ,in, is the sample size of the first set of sample data, that is, the number of trials of the first set of sample data, is the sample size of the second set of sample data, that is, the number of trials of the second set of sample data, The sample size The corresponding number of successes, The sample size The first set of sample data , the second set of sample data The corresponding success probability , They are: ; .

[0039] against Figure 1 In the technical solution shown, in some possible implementations, the thresholds of the test bias, the risk of false positives and the risk of false positives of the two groups of sample data are set, and the risk ratio is obtained according to the thresholds of the risk of false positives and the risk of false positives. Specifically, when performing consistency tests on two groups of sample data that follow a binomial distribution, three important concepts of test bias, risk of false positives and risk of false positives are usually involved. In consistency tests, it is usually hoped to control the probabilities of the risk of false positives and the risk of false positives and to strike a balance between the two to ensure correct statistical inferences. In practical applications, the acceptable range of test bias, the size of the risk of false positives and the level of false positives that are expected to be achieved can be set according to the research questions and specific circumstances to ensure the reliability and accuracy of hypothesis testing. Among them, the test bias is the difference between the estimated probability and the true probability. If there is a systematic deviation between the estimated probability and the true probability, that is, the estimated probability always deviates from the true probability in a fixed direction in repeated sampling, it is called a deviation, and the existence of the test bias may lead to inaccurate estimation results. The test bias is usually expressed as Indicates that in order to make the estimated probability as close to the true probability as possible, the setting rule is as small as possible. The size of the test bias can be evaluated by comparing the difference between the estimated probability and the true probability. The risk of rejecting the truth, also known as the risk of the first type of error, refers to the possibility of rejecting a true hypothesis. In other words, the risk of rejecting the truth refers to the probability of mistakenly rejecting the null hypothesis when it is true. The risk of rejecting the truth is usually represented by α, which represents the upper limit of the allowable risk of rejecting the truth. The smaller the risk of rejecting the truth α, the better. Common α values ​​are 0.01 and 0.05. The risk of false positives, also known as the risk of the second type of error, refers to the possibility of accepting a false hypothesis. In other words, the risk of false positives refers to the probability of mistakenly accepting the null hypothesis when the alternative hypothesis is true, or the probability of accepting the null hypothesis but the null hypothesis is actually not true. The risk of false positives is usually expressed as It is usually related to statistical power, where the statistical power refers to the probability of rejecting a false hypothesis or the probability of rejecting the null hypothesis when the alternative hypothesis is true, that is, , common values ​​are 0.8 and 0.9. Usually, we hope that the risk of false positives is as small as possible, that is, we hope that the efficacy is as high as possible.

[0040] It should be noted that in statistics, the null hypothesis and the alternative hypothesis are two important concepts in hypothesis testing, which are used to make hypothetical inferences and tests on research questions. Among them, the null hypothesis is an assumption about the population parameter. The null hypothesis usually indicates that there is no effect, no association or no difference, etc., and is a conservative assumption. In hypothesis testing, data is collected to determine whether the null hypothesis should be rejected. The alternative hypothesis is a hypothesis opposite to the null hypothesis, which usually expresses the existence of an effect, association or difference. In hypothesis testing, rejecting the null hypothesis usually means accepting the alternative hypothesis.

[0041] For the above implementation, in some examples, the risk ratio is obtained according to the thresholds of the true rejection risk and the false existence risk. Specifically, the risk ratio is calculated according to the set thresholds of the true rejection risk and the false existence risk as follows:

[0042] in, r is the risk ratio, α To avoid real risks, β To maintain the risk of counterfeiting.

[0043] It should be noted that in statistical tests, the optimal test is not the minimum of a single risk of α or β, but Minimize, where a and b are coefficients, and usually a+b=1 should be satisfied. In other words, there is no way to minimize both the risk of rejecting the true answer and the risk of surviving the false answer. Usually, reducing the risk of one error comes at the expense of increasing the risk of another error. The risk ratio is used to quantify the trade-off between the risk of rejecting the true answer and the risk of surviving the false answer, and the comparison result between the risk of rejecting the true answer and the risk of surviving the false answer can be intuitively known. The closer the risk ratio is to 1, the relative balance between the risk of rejecting the true answer and the risk of surviving the false answer. If the risk ratio is greater than 1, it means that the risk of surviving the false answer is high, and the experimental design or hypothesis test setting needs to be reconsidered.

[0044] against Figure 1 In some possible implementations of the technical solution shown, the determination of the test parameter value according to the comparison result of the sample size in the two groups of sample data and the success probability includes:

[0045] Determine the distribution parameter value of the null hypothesis, the total number of trials, and the total number of successes parameter value according to the first comparison result of the sample sizes of the two groups of sample data;

[0046] The distribution parameter value of the alternative hypothesis is determined according to the second comparison result of the success probabilities of the two groups of sample data, the distribution parameter value of the null hypothesis and the threshold value of the test deviation.

[0047] For the above implementation, specifically, the test parameter value includes at least the distribution parameter value of the null hypothesis, the distribution parameter value of the alternative hypothesis, the total number of trials, and the total number of successes parameter value. The null hypothesis and the alternative hypothesis are the basic concepts of hypothesis testing in statistics, wherein the null hypothesis usually represents the hypothesis that there is no effect or difference, and the distribution parameter value of the null hypothesis is at a specific value or is located in a certain interval. The alternative hypothesis represents the hypothesis that there is an effect or difference, and the distribution parameter value of the alternative hypothesis is a different value or is located in other spaces in the parameter space. The parameter space is the range of possible values ​​of the distribution parameter. First, the null hypothesis and the alternative hypothesis of the hypothesis test are given:

[0048]

[0049] in, is the original hypothesis, For the alternative hypothesis, is the binomial distribution parameter, is the distribution parameter value of the null hypothesis, is the distribution parameter value for the alternative hypothesis.

[0050] Combined with the first set of sample data , the second set of sample data In some examples, the distribution parameter value of the null hypothesis, the total number of trials, and the total number of success parameters are determined according to the first comparison result of the sample sizes of the two groups of sample data. Specifically, there are two cases. The first case is:

[0051] First, compare the sample sizes in the two sets of sample data;

[0052] if ,but , , ;

[0053] Then, compare the success probabilities corresponding to the two sets of sample data and size;

[0054] if ,but ,otherwise ;

[0055] The second case is:

[0056] First, compare the sample sizes in the two sets of sample data;

[0057] if ,but , , ;

[0058] Then, compare the success probabilities corresponding to the two sets of sample data and size;

[0059] if ,but ,otherwise .

[0060] in, is the number of trials for the first set of sample data, is the number of trials for the second set of sample data, The sample size The corresponding number of successes, for The corresponding number of successes, is the distribution parameter value of the null hypothesis, n is the total number of trials, k is the total number of successes, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, To test the deviation.

[0061] against Figure 1 In some possible implementations of the technical solution shown, the step of calculating the test threshold value according to the risk ratio and the test parameter value includes:

[0062] The likelihood ratio is calculated using the likelihood function of the binomial distribution:

[0063]

[0064] in, is the likelihood ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, is the observed data, n is the number of observation data;

[0065] The test threshold is obtained according to the likelihood ratio and risk ratio:

[0066]

[0067] in, c is the test threshold, is the risk ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, n is the number of observation data.

[0068] For the above implementation, specifically, for the probability index based on binomial distribution, the distribution parameters are tested through simulation experiments. , that is, whether the binomial distribution probability p meets the design requirements, it is necessary to set the acceptance domain and rejection domain.

[0069] First, give the null hypothesis and alternative hypothesis of the hypothesis test:

[0070]

[0071] in, is the original hypothesis, For the alternative hypothesis, is the binomial distribution parameter, is the distribution parameter value of the null hypothesis, is the distribution parameter value for the alternative hypothesis.

[0072] In hypothesis testing, the null hypothesis usually indicates that there is no significant difference between the two sets of data, while the alternative hypothesis indicates that there is a significant difference between the two sets of data. The acceptance domain and the rejection domain are two concepts used to determine whether to reject the null hypothesis. The acceptance domain refers to a set of possible sample observations. If the sample observations fall within the acceptance domain, the null hypothesis is accepted; the rejection domain refers to another set of possible sample observations. If the sample observations fall within the rejection domain, the null hypothesis is rejected. Combining the aforementioned null hypothesis and alternative hypothesis, the risk of false rejection refers to the probability of incorrectly rejecting the null hypothesis when the null hypothesis is true, which can be expressed as: The falsification risk refers to the probability of incorrectly accepting the null hypothesis when the alternative hypothesis is true, which can be expressed as: .

[0073] In some examples, the likelihood ratio is obtained by calculating the likelihood function of the binomial distribution. Specifically, the optimal test is not the minimum of a single risk of α or β, but Minimize, where a and b are coefficients, and usually should satisfy a+b=1. The likelihood function Indicates that given parameters Under, the observed data The likelihood function is the product of the probability density function or probability mass function of the known data. In the case of The likelihood ratio is the ratio of the maximum value of the likelihood function under the alternative hypothesis to the maximum value of the likelihood function under the null hypothesis. The likelihood ratio test is a statistical method used to compare the relative goodness of fit of two competing models. In detail, the likelihood function Defined as:

[0074]

[0075] in, is the binomial distribution parameter, is the observed data, For a given Time observation The probability density function of .

[0076] The hypothesis test can be expressed as:

[0077] when If ;

[0078] when If .

[0079] Defining the likelihood ratio ,make , then the hypothesis test is further transformed into:

[0080]

[0081] In some examples, obtaining a test threshold according to the likelihood ratio and the risk ratio includes:

[0082] Define the likelihood function for the binomial distribution for:

[0083]

[0084] The likelihood ratio is calculated based on the likelihood function of the binomial distribution:

[0085]

[0086] Taking the logarithm of both sides of the equation yields:

[0087]

[0088] According to the relationship between likelihood ratio and risk ratio , convert the equation into:

[0089]

[0090] Introducing an intermediate variable, the test threshold c is:

[0091]

[0092] in, c is the test threshold, is the risk ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, n is the number of observation data.

[0093] It should be noted that the test threshold is a threshold calculated based on the set test deviation, true rejection risk and false positive risk and the determined test parameter value, which is used to determine the boundaries of the rejection region and the acceptance region. If the test statistic is less than or equal to the test threshold, the null hypothesis is accepted and the two sets of data are considered to be consistent. If the test statistic is greater than the test threshold, the null hypothesis is rejected and the two sets of data are considered to be inconsistent.

[0094] against Figure 1 In some possible implementations of the technical solution shown, the comparing the inspection threshold with the inspection parameter value to determine the consistency of the two sets of sample data includes:

[0095] When the distribution parameter of the null hypothesis is less than the distribution parameter of the alternative hypothesis and the total number of tests is less than the test threshold, the two sets of sample data pass the consistency check; or,

[0096] When the distribution parameter of the null hypothesis is greater than or equal to the distribution parameter of the alternative hypothesis and the total number of tests is greater than or equal to the test threshold, the two sets of sample data pass the consistency check.

[0097] For the above implementation, specifically, the distribution parameters of the null hypothesis and the distribution parameters of the alternative hypothesis are compared to obtain a comparison result; when ,if If the consistency check is successful, it passes the consistency check, otherwise it fails. ,if If yes, the consistency check is passed; otherwise, it fails.

[0098] Combination Figure 1 The technical solution shown in the figure is described in detail with reference to the technical solution of the method. Figure 2 , which shows a detailed flow chart of a method for checking consistency of multi-source probability test data based on binomial distribution provided by an embodiment of the present disclosure. Specifically, two groups of sample data of typical probability indicators based on binomial distribution are selected. , , where the first set of sample data The data value is (800, 600), which can represent a set of digital simulation test results. The second set of sample data The data value of is (200, 146), which can represent a set of semi-physical simulation test results. The first set of sample data and the second set of sample data both obey the binomial distribution. Figure 1 The technical solution shown checks the consistency of two sets of sample data. The detailed operation steps are as follows:

[0099] S201: Obtain sample data of two groups of indicators that obey binomial distribution probability , , the corresponding data values ​​are (800, 600) and (200, 146);

[0100] S202: Calculate the success probabilities corresponding to the two groups of sample data respectively;

[0101] The probability corresponding to the first set of sample data is:

[0102] The probability corresponding to the second set of sample data is:

[0103] S203: Setting the inspection deviation δ , abandon the real risk α and the risk of counterfeiting β threshold and calculate the hazard ratio;

[0104] Setting the test deviation , abandon the real risk , there is a risk of counterfeiting , according to the risk of abandonment α and the risk of counterfeiting β Calculate the hazard ratio: .

[0105] S204: Compare the sample sizes of two groups of sample data , ;

[0106] S205: Judgment Is it greater than or equal to If yes, jump to step S206, otherwise, jump to step S207;

[0107] S206: If , determine the test parameter value in the first case;

[0108] Comparing the sample sizes of two groups of data ,but , , .

[0109] S207: If , determine the test parameter value in the second case;

[0110] The test parameter values ​​for the second case are: , , .

[0111] It should be noted that steps S206 and S207 respectively determine corresponding test parameter values ​​according to the comparison results of the sample sizes of the two groups of sample data, and the two are in an either-or relationship.

[0112] S208: Judgment Is it greater than or equal to , if yes, jump to step S209, otherwise, jump to step S210;

[0113] S209: When and , determine the distribution parameters of the alternative hypothesis in the first case;

[0114] because and ,but .

[0115] S210: When and , determine the distribution parameters of the alternative hypothesis in the second case;

[0116] when and ,but .

[0117] S211: Calculating a test threshold c according to the test parameter value and the risk ratio;

[0118] The test threshold is: .

[0119] S212: consistency check.

[0120] because and , then the consistency check is passed.

[0121] In summary, the consistency test results of the two groups of sample data based on binomial distribution are: typical probability index data (800, 600) representing a group of digital simulation test results and typical probability index data (200, 146) representing a group of semi-physical simulation test results. , abandon the real risk , there is a risk of counterfeiting The consistency test is passed under the condition of

[0122] Based on the same inventive concept as the above technical solution, see Figure 3 , which shows a multi-source probability test data consistency test device 300 based on binomial distribution, the device 300 includes: a calculation part 301, a setting part 302, a determination part 303, an acquisition part 304 and a judgment part 305; wherein,

[0123] The calculation part 301 is configured to calculate the corresponding success probabilities for the two groups of sample data that obey the binomial distribution according to the sample size and the number of successes;

[0124] The setting part 302 is configured to set the thresholds of the inspection deviation, the risk of false positives and the risk of false positives of the two groups of sample data, and obtain the risk ratio according to the thresholds of the risk of false positives and the risk of false positives;

[0125] The determining part 303 is configured to determine the inspection parameter value according to the sample size in the two sets of sample data and the comparison result of the success probability;

[0126] The obtaining part 304 is configured to calculate and obtain a test threshold value according to the risk ratio and the test parameter value;

[0127] The determination part 305 is configured to compare the inspection threshold with the inspection parameter value to determine the consistency of the two sets of sample data.

[0128] In some examples, the computing portion 301 is configured to:

[0129] Divide the number of successes of the first set of sample data by the sample size to obtain the corresponding probability of success of the first set of sample data;

[0130] The number of successes of the second set of sample data is divided by the sample size to obtain the corresponding probability of success of the second set of sample data.

[0131] In some examples, the setting portion 302 is configured to:

[0132] The risk ratio is calculated based on the set threshold of the risk of true rejection and the threshold of the risk of false existence: ;

[0133] in, r is the risk ratio, α To avoid real risks, β To maintain the risk of counterfeiting.

[0134] In some examples, the determining portion 303 is configured to:

[0135] Determine the distribution parameter value of the null hypothesis, the total number of trials, and the total number of successes parameter value according to the first comparison result of the sample sizes of the two groups of sample data;

[0136] The distribution parameter value of the alternative hypothesis is determined according to the second comparison result of the success probabilities of the two groups of sample data, the distribution parameter value of the null hypothesis and the threshold value of the test deviation.

[0137] In some examples, the obtaining portion 304 is configured to:

[0138] The likelihood ratio is calculated using the likelihood function of the binomial distribution:

[0139]

[0140] in, is the likelihood ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, is the observed data, n is the number of observation data;

[0141] The test threshold is obtained according to the likelihood ratio and risk ratio:

[0142]

[0143] in, c is the test threshold, is the risk ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, n is the number of observation data.

[0144] In some examples, the determining portion 305 is configured to:

[0145] When the distribution parameter of the null hypothesis is less than the distribution parameter of the alternative hypothesis and the total number of tests is less than the test threshold, the two sets of sample data pass the consistency check; or,

[0146] When the distribution parameter of the null hypothesis is greater than or equal to the distribution parameter of the alternative hypothesis and the total number of tests is greater than or equal to the test threshold, the two sets of sample data pass the consistency check.

[0147] It can be understood that the exemplary technical solution of the multi-source probability test data consistency verification device 300 based on binomial distribution belongs to the same concept as the technical solution of the multi-source probability test data consistency verification method based on binomial distribution. Therefore, the details not described in detail in the technical solution of the multi-source probability test data consistency verification device 300 based on binomial distribution can be referred to the description of the technical solution of the multi-source probability test data consistency verification method based on binomial distribution. The disclosed embodiment will not elaborate on this.

[0148] refer to Figure 4 , which shows a schematic diagram of the hardware structure of a computing device provided by an exemplary embodiment of the present disclosure. In some examples, the computing device may be at least one of a smart phone, a smart watch, a desktop computer, a laptop, a virtual reality terminal, an augmented reality terminal, a wireless terminal, and a laptop portable computer. The computing device has a communication function and can access a wired network or a wireless network. The computing device may generally refer to one of a plurality of terminals, and those skilled in the art may know that the number of the above terminals may be more or less. In some examples, the computing device may receive data for consistency check of multi-source probability test data based on a binomial distribution based on the wired network or wireless network to which it is connected. It can be understood that the computing device undertakes the calculation and processing work of the technical solution of the present disclosure, and the embodiments of the present disclosure are not limited to this.

[0149] like Figure 4 As shown, the computing device in the present disclosure may include one or more of the following components: a processor 410 and a memory 420 .

[0150] Optionally, the processor 410 uses various interfaces and lines to connect various parts within the entire computing device, and executes various functions of the computing device and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 420, and calling data stored in the memory 420. Optionally, the processor 410 can be implemented in at least one hardware form of digital signal processing (DSP), field programmable gate array (FPGA), and programmable logic array (PLA). The processor 410 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), a neural network processor (NPU), and a baseband chip. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing the content that needs to be displayed on the touch display; the NPU is used to implement artificial intelligence (AI) functions; and the baseband chip is used to process wireless communications. It is understandable that the above baseband chip may not be integrated into the processor 410, but may be implemented by a separate chip.

[0151] The memory 420 may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory 420 includes a non-transitory computer-readable storage medium. The memory 420 may be used to store instructions, programs, codes, code sets, or instruction sets. The memory 420 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above various method embodiments, etc.; the data storage area may store data created according to the use of the computing device, etc.

[0152] In addition, those skilled in the art can understand that the structure of the computing device shown in the above drawings does not constitute a limitation on the computing device, and the computing device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. For example, the computing device also includes a display screen, a camera assembly, a microphone, a speaker, a radio frequency circuit, an input unit, a sensor (such as an acceleration sensor, an angular velocity sensor, an optical fiber sensor, etc.), an audio circuit, a WiFi module, a power supply, a Bluetooth module and other components, which will not be described in detail here.

[0153] An embodiment of the present disclosure also provides a computer-readable storage medium storing at least one instruction, wherein the at least one instruction is used to be executed by a processor to implement the multi-source probability test data consistency verification method based on binomial distribution as described in the above embodiments.

[0154] The embodiments of the present disclosure also provide a computer program product, which includes computer instructions, which are stored in a computer-readable storage medium; a processor of a computing device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computing device executes to implement the multi-source probability test data consistency verification method based on binomial distribution described in the above-mentioned embodiments.

[0155] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the embodiments of the present disclosure can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any media that facilitates the transmission of computer programs from one place to another. Storage media can be any available media that can be accessed by general or special-purpose computers.

[0156] The above is only a specific embodiment of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present disclosure, which should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

Claims

1. A method for checking consistency of multi-source probability test data based on binomial distribution, characterized in that: The method comprises: For the two groups of sample data used to characterize the digital simulation test results and the semi-physical simulation test results in the probability indicators of missiles that obey binomial distribution, the corresponding success probabilities are calculated according to the sample size and the number of successes. Setting thresholds of test deviation, risk of rejecting a true result, and risk of retaining a false result for the two groups of sample data, and obtaining a risk ratio according to the thresholds of risk of rejecting a true result and risk of retaining a false result; Determining a test parameter value according to a comparison result of the sample size in the two groups of sample data and the success probability; Calculate and obtain a test threshold value according to the risk ratio and the test parameter value; Comparing the inspection threshold with the inspection parameter value to determine the consistency of the two sets of sample data; The step of setting the thresholds of the inspection deviation, the risk of rejecting a true result, and the risk of false positives for the two groups of sample data, and obtaining the risk ratio according to the thresholds of the risk of rejecting a true result and the risk of false positives, includes: The risk ratio is calculated based on the set threshold of the risk of true rejection and the threshold of the risk of false existence: ; in, r is the risk ratio, α To avoid real risks, β To avoid the risk of counterfeiting; in, , ; in, is the original hypothesis, For the alternative hypothesis, is the binomial distribution parameter, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis; The step of calculating the test threshold value according to the risk ratio and the test parameter value comprises: The likelihood ratio is calculated using the likelihood function of the binomial distribution: in, is the likelihood ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, is the observed data, n is the number of observation data; The test threshold is obtained according to the likelihood ratio and risk ratio: in, c is the test threshold, is the risk ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, n is the number of observation data; The comparing the inspection threshold with the inspection parameter value to determine the consistency of the two sets of sample data includes: When the distribution parameter of the null hypothesis is less than the distribution parameter of the alternative hypothesis and the total number of tests is less than the test threshold, the two sets of sample data pass the consistency check; or, When the distribution parameter of the null hypothesis is greater than or equal to the distribution parameter of the alternative hypothesis and the total number of tests is greater than or equal to the test threshold, the two sets of sample data pass the consistency check.

2. The method according to claim 1, characterized in that: The two groups of sample data that obey the binomial distribution are respectively calculated to obtain corresponding success probabilities according to the sample size and the number of successes, including: Divide the number of successes of the first set of sample data by the sample size to obtain the corresponding probability of success of the first set of sample data; The number of successes of the second set of sample data is divided by the sample size to obtain the corresponding probability of success of the second set of sample data.

3. The method according to claim 1, characterized in that: The step of determining the test parameter value according to the comparison result of the sample size in the two groups of sample data and the success probability comprises: Determine the distribution parameter value of the null hypothesis, the total number of trials, and the total number of successes parameter value according to the first comparison result of the sample sizes of the two groups of sample data; The distribution parameter value of the alternative hypothesis is determined according to the second comparison result of the success probabilities of the two groups of sample data, the distribution parameter value of the null hypothesis and the threshold value of the test deviation.

4. A device for consistency testing of multi-source probability test data based on binomial distribution, characterized in that: The device comprises: a calculation part, a setting part, a determination part, an acquisition part and a judgment part; wherein, The calculation part is configured to calculate the corresponding success probability according to the sample size and the number of successes for two groups of sample data used to characterize the digital simulation test results and the semi-physical simulation test results in the probability index of the missile obeying the binomial distribution; The setting part is configured to set the thresholds of the inspection deviation, the risk of rejecting a true false positive and the risk of excluding a false positive of the two groups of sample data, and obtain the risk ratio according to the thresholds of the risk of rejecting a true false positive and the risk of excluding a false positive; The determining part is configured to determine the inspection parameter value according to the sample size in the two sets of sample data and the comparison result of the success probability; The obtaining part is configured to calculate and obtain a test threshold value according to the risk ratio and the test parameter value; The judgment part is configured to compare the inspection threshold with the inspection parameter value to judge the consistency of the two sets of sample data; The step of setting the thresholds of the inspection deviation, the risk of rejecting a true result, and the risk of false positives for the two groups of sample data, and obtaining the risk ratio according to the thresholds of the risk of rejecting a true result and the risk of false positives, includes: The risk ratio is calculated based on the set threshold of the risk of true rejection and the threshold of the risk of false existence: ; in, r is the risk ratio, α To avoid real risks, β To avoid the risk of counterfeiting; in, , ; in, is the original hypothesis, For the alternative hypothesis, is the binomial distribution parameter, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis; The step of calculating the test threshold value according to the risk ratio and the test parameter value comprises: The likelihood ratio is calculated using the likelihood function of the binomial distribution: in, is the likelihood ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, is the observed data, n is the number of observation data; The test threshold is obtained according to the likelihood ratio and risk ratio: in, c is the test threshold, is the risk ratio, is the distribution parameter value of the null hypothesis, is the distribution parameter value of the alternative hypothesis, n is the number of observation data; The comparing the inspection threshold with the inspection parameter value to determine the consistency of the two sets of sample data includes: When the distribution parameter of the null hypothesis is less than the distribution parameter of the alternative hypothesis and the total number of tests is less than the test threshold, the two sets of sample data pass the consistency check; or, When the distribution parameter of the null hypothesis is greater than or equal to the distribution parameter of the alternative hypothesis and the total number of tests is greater than or equal to the test threshold, the two sets of sample data pass the consistency check.

5. The device according to claim 4, characterized in that: The calculation part is configured as follows: Divide the number of successes of the first set of sample data by the sample size to obtain the corresponding probability of success of the first set of sample data; The number of successes of the second set of sample data is divided by the sample size to obtain the corresponding probability of success of the second set of sample data.

6. A computing device, characterized in that The computing device comprises: a processor and a memory; wherein, The memory is used to store a computer program that can be run on the processor; The processor is used to execute the multi-source probability test data consistency verification method based on binomial distribution as described in any one of claims 1 to 3 when running the computer program.

7. A computer storage medium, characterized in that: The storage medium stores at least one instruction, and the at least one instruction is used to be executed by a processor to implement the multi-source probability test data consistency verification method based on binomial distribution as claimed in any one of claims 1 to 3.

Images