An accident economic loss influence factor analysis method and system

Abstract

The present application relates to a kind of accident economic loss influence factor analysis method and system, in method part, first based on GB2 model, establish multiple stochastic parameter model, then using LOOIC value carries out effect comparison, selects suitable stochastic parameter model and obtains the positive influence or negative influence of various explanatory variables to accident economic loss;Finally, based on sensitivity analysis, the influence degree of explanatory variable to accident economic loss is excavated.Through the scheme of the present application, an objective, comprehensive, accurate analysis method for excavating accident economic loss influence factor is provided, and the application range is wider, and the influence of related factors on the economic loss caused by accident is better represented.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data science and technology, and in particular, relates to an accident economic loss influencing factor analysis method and system. BACKGROUND

[0002] Unlike the severity of accidents, the economic loss of fishing vessel accidents is continuous, and the analysis result of the economic loss of fishing vessel accidents provides detailed information of the loss of human life and the loss of ship property. Therefore, it is crucial to analyze the influence of influencing factors on the economic loss caused by fishing vessel accidents through technical means.

[0003] The economic loss of ship accidents is a non-negative continuous variable, and the distribution of such catastrophic economic loss deviates from normality, often showing right skewness, excessive kurtosis and heavy tail characteristics. In the prior art, Tobit regression method, generalized F distribution model, Bayesian lognormal regression model, etc. are used to explain the non-negativity of property loss and total loss, but these technical solutions largely ignore the problem of heavy-tailed distribution, thereby showing deficiencies in accurately describing the complex distribution characteristics of economic loss and lacking flexibility in heavy-tailed distribution modeling.

[0004] In addition, the occurrence of fishing vessel accidents is extremely complex and involves numerous factors, and it is impossible to cover all safety factors. Unobserved safety factors can have different effects on variables related to these unobserved factors, i.e. unobserved heterogeneity problem. However, most of the existing models constructed by the analysis method of fishing vessel accidents assume that the distribution mean of each random parameter remains unchanged between observations, ignoring the bias of the estimated parameters caused by unobserved heterogeneity, which can cause model specification errors, leading to inaccurate inferences and possible underestimation or overestimation of direct marginal effects. SUMMARY

[0005] The technical problem to be solved by the present application is how to solve the problem of analyzing influencing factors when the economic loss of ship accidents shows heavy-tailed distribution and the problem of ignoring the bias of estimated parameters caused by unobserved heterogeneity in the analysis method of the prior art. In order to overcome the defects of the prior art, the present application provides an accident economic loss influencing factor analysis method and an accident economic loss influencing factor analysis system.

[0006] The accident economic loss influencing factor analysis method provided by the present application comprises the following steps:

[0007] S1: constructing a Bayesian random parameter GB2 model according to the observed information of the sample characteristics of the obtained fishing vessel accidents;

[0008] S2: constructing a plurality of random parameter models for analyzing the economic loss influencing factors of the fishing vessel accident based on the Bayesian random parameter GB2 model and the observation information of the sample characteristics;

[0009] S3: calculating the leave-one-out cross-validation information criterion value of each random parameter model by the calculation module;

[0010] S4: selecting the random parameter model with a lower leave-one-out cross-validation information criterion value to analyze the economic loss influencing factors of the fishing vessel accident, and obtaining the positive or negative influence of various explanatory variables on the economic loss of the fishing vessel accident;

[0011] S5: based on the positive or negative influence of various explanatory variables on the economic loss of the fishing vessel accident obtained in the step S4, the influence degree of the explanatory variables on the economic loss of the accident is excavated.

[0012] The disclosed accident economic loss influencing factor analysis method aims to solve the above problems. Firstly, the Bayesian random parameter GB2 model is constructed through observation information. The model can solve the problem of highly non-normal distribution and heavy-tailed distribution. Then, based on the model, a plurality of random parameter models are established, which can improve the statistical fitting degree and provide more insight. Then, the leave-one-out cross-validation information criterion value is used for effect comparison, and the appropriate random parameter model is selected for analysis, and the positive or negative influence of various explanatory variables on the economic loss of the accident is obtained. Finally, based on the sensitivity analysis, the influence degree of the explanatory variables on the economic loss of the accident is excavated, thereby overcoming the problems in the prior art, such as the difficulty in analyzing the influencing factors when the economic loss of the ship accident is heavy-tailed distribution, and the neglect of unobserved heterogeneity to cause bias in the estimated parameters. The method is objective, comprehensive and accurate, can excavate the influencing factors of the economic loss of the accident, has a wide application range, and better represents the influence of related factors on the economic loss caused by the accident.

[0013] In one possible implementation, in the Bayesian random parameter GB2 model:

[0014] The probability density function of the random variable is as follows:

[0015]

[0016] In the formula,

[0017] a is a shape parameter, which determines the peak value and tail thickness of the economic loss distribution;

[0018] p is the probability of the accident with small economic loss;

[0019] q is the probability of the accident with major economic loss;

[0020] B(p, q) is a B function;

[0021] λ reflects the economic loss level of the accident, is a scale parameter, controls the overall level and range of loss, and its calculation formula is as follows:

[0022] λ = exp(β'x) = exp(β0+β1x1+β2x2+...+βkxk), k x k ),

[0023] In the formula,

[0024] x is a covariate vector affecting the economic loss of the fishing boat accident;

[0025] β is a coefficient vector corresponding to x;

[0026] k is the number of covariates;

[0027] The k-th moment of the random variable is as follows:

[0028]

[0029] This scheme can realize the association of the distribution mean and variance of the random parameter with the explanatory variable, and there is a difference between different observation values, so as to further improve the model performance, and effectively solve the problems of highly non-normal distribution and heavy-tailed distribution.

[0030] In a possible implementation, the plurality of random parameter models in the step S2 include a general random parameter model, a mean heterogeneity random parameter model, and a mean and variance heterogeneity random parameter model; and further help to explain the unobserved heterogeneity caused by unobserved safety factors, so as to improve the statistical fitting degree and provide more insights.

[0031] In a possible implementation, the general random parameter model is expressed as follows:

[0032]

[0033] In the formula,

[0034] β n is a random parameter vector;

[0035] b is an average parameter vector of all observation values of the sample characteristics of the fishing boat accident;

[0036] is an error term of a random distribution, used to capture possible heterogeneity.

[0037] In a possible implementation, the mean heterogeneity random parameter model is expressed as follows:

[0038]

[0039] In the formula,

[0040] Z n To affect β n The vector of explanatory variables for the mean;

[0041] Θ is Z n The corresponding estimable parameter vector; if the elements in Θ are significantly different from zero, it indicates that the mean heterogeneous random parameter model has mean heterogeneity.

[0042] In one possible implementation, the mean and variance variability random parameter model is expressed as follows:

[0043]

[0044] In the formula,

[0045] W n This is a vector of explanatory variables used to capture the heterogeneity of the standard deviation σ;

[0046] ω is W n The corresponding estimable parameter vector;

[0047] If no element in ω is significantly different from zero, it means that the mean and variance heterogeneity random parameter model only has the heterogeneity of the mean.

[0048] If no element in Θ is significantly different from zero, it means that the mean and variance qualitative random parameter model degenerates into the general random parameter model.

[0049] In one possible implementation, in step S3, the calculation module obtains the following formula for calculating the leave-one-out cross-validation information criterion value:

[0050] LOOIC = -2ELPD LOO ,

[0051]

[0052] In the formula,

[0053] p M (y i |θ) represents the M probability in the regression model;

[0054] p M (θ|y -i ) represents the posterior value of parameter θ, where the observed value y i Excluded;

[0055] The scheme provides a calculation formula of leave-one-out cross-validation information criterion value, ensures to perform effect comparison, thereby selects a suitable random parameter model for analysis, and obtains positive or negative influence of various explanatory variables on the economic loss of the accident.

[0056] Another technical solution of the present application is to provide an accident economic loss influencing factor analysis system, comprising a circuit connection relationship:

[0057] A model module is configured to construct a Bayesian random parameter GB2 model according to observation information of sample characteristics of the obtained fishing vessel accident;

[0058] A parameter model module is configured to construct a plurality of random parameter models for analyzing the economic loss influencing factors of the fishing vessel accident based on the Bayesian random parameter GB2 model and the observation information of the sample characteristics;

[0059] A calculation module is configured to calculate a LOOIC value of each random parameter model respectively;

[0060] A selection module is configured to select the random parameter model with a lower leave-one-out cross-validation information criterion value for fishing vessel accident economic loss influencing factor analysis, and obtain positive or negative influence of various explanatory variables on the economic loss of the fishing vessel accident;

[0061] An analysis module is configured to obtain the influence degree of the explanatory variables on the economic loss of the accident.

[0062] The disclosed accident economic loss influencing factor analysis system solves the above problems by setting a model module to construct a Bayesian random parameter GB2 model from observation information, which can solve the problem of highly non-normal distribution and heavy-tailed distribution; on this basis, a parameter model module is set to establish a plurality of random parameter models based on the model, which can improve statistical fitting degree and provide more insight; through the cooperation of the calculation module and the selection module, the leave-one-out cross-validation information criterion value is used for effect comparison, a suitable random parameter model is selected for analysis, and positive or negative influence of various explanatory variables on the economic loss of the accident is obtained; finally, the analysis module is based on sensitivity analysis to mine the influence degree of the explanatory variables on the economic loss of the accident, thereby overcoming the problems of the existing analysis method, such as difficulty in handling the influencing factor analysis of the fishing vessel accident economic loss with heavy-tailed distribution, and ignoring the unobserved heterogeneity to cause bias in the estimated parameters, and other problems, which is objective, comprehensive and accurate, can mine the influencing factors of the accident economic loss, has a wide application range, and better represents the influence of related factors on the economic loss caused by the accident. BRIEF DESCRIPTION OF DRAWINGS

[0063] Figure 1 A flow chart of an accident economic loss influencing factor analysis method is disclosed in the embodiments of the present application;

[0064] Figure 2 A structural schematic diagram of an accident economic loss influencing factor analysis system is disclosed in the embodiments of the present application. DETAILED DESCRIPTION

[0065] First, those skilled in the art should understand that these embodiments are only used to explain the technical principles of the embodiments of the present application, and are not intended to limit the protection scope of the embodiments of the present application. Those skilled in the art can adjust them as needed in order to adapt to specific application occasions.

[0066] In the description of the embodiments of the present application, it should be noted that, unless otherwise explicitly specified and limited, the terms "electrical connection", "electrical connection relationship" should be understood in a broad sense, i.e. to refer to the connection mode with electrical relationship, for example, it can be through a wire to realize circuit connection, or through a wireless signal channel (channel) to realize electrical connection, or a combination of the two. In addition, "electrical connection", "electrical connection relationship" can be established on the basis of mechanical connection (such as the wire is arranged in the connection key); it can be directly connected, or indirectly connected through an intermediate medium. For those skilled in the art, the specific meaning of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0067] The present application will be further described in detail below by adopting two embodiments and combining with the drawings and specific embodiments.

[0068] Embodiment one:

[0069] Referring to Figure 1 It is disclosed that the embodiments of the present application disclose an accident economic loss influencing factor analysis method, which comprises the following steps:

[0070] S1: Constructing a Bayesian random parameter GB2 model according to the observed information of the sample characteristics of the obtained fishing boat accident.

[0071] In this embodiment, the economic loss of the fishing boat accident is taken as the response variable, the distribution parameter is specified as the function of the covariate, and then the relationship between the accident economic loss data and the explanatory variable is mined. On this basis, it is assumed that the scale parameter λ can change with the covariate, and then in the obtained Bayesian random parameter GB2 model: the probability density function of the random variable is as follows:

[0072]

[0073] In the formula,

[0074] a is a shape parameter, which determines the peak and tail thickness of the economic loss distribution;

[0075] p is the probability of an accident with a small economic loss;

[0076] q is the probability of an accident with a major economic loss;

[0077] B(p, q) is a B function;

[0078] λ reflects the economic loss level of the accident, which is a scale parameter, and controls the overall level and range of the loss, and is calculated as follows:

[0079] λ = exp(β'x) = exp(β0+β1x1+β2x2+...+βkxk), k x k ),

[0080] wherein,

[0081] x is a covariate vector that affects the economic loss of the fishing vessel accident;

[0082] β is a coefficient vector corresponding to x;

[0083] k is the number of covariates.

[0084] At the same time, in the Bayesian random parameter GB2 model, the kth moment of the random variable is as follows:

[0085]

[0086] S2: Constructing multiple random parameter models for analyzing the influencing factors of the economic loss of the fishing vessel accident based on the Bayesian random parameter GB2 model and the observed information of the sample characteristics.

[0087] In this embodiment, the multiple random parameter models in step S2 include a general random parameter model, a mean heterogeneity random parameter model, and a mean and variance heterogeneity random parameter model, wherein the general random parameter model (hereinafter referred to as RPM) is expressed as follows:

[0088]

[0089] wherein,

[0090] β n is a random parameter vector;

[0091] b is a mean parameter vector of all observed values of the sample characteristics of the fishing vessel accident;

[0092] is an error term of a random distribution, which is used to capture possible heterogeneity.

[0093] The mean heterogeneous stochastic parametric model (hereinafter referred to as RPMHM) is described as follows:

[0094]

[0095] In the formula,

[0096] Z n To affect β n The vector of explanatory variables for the mean;

[0097] Θ is Z n The corresponding estimable parameter vector; if the elements in Θ are significantly different from zero, it indicates that the mean heterogeneous random parameter model has mean heterogeneity.

[0098] The mean-variance qualitative random parametric model (hereinafter referred to as RPMHMV) is expressed as follows:

[0099]

[0100] In the formula,

[0101] W n This is a vector of explanatory variables used to capture the heterogeneity of the standard deviation σ;

[0102] ω is W n The corresponding estimable parameter vector; if no element in ω is significantly different from zero, it means that the mean and square heterogeneous random parameter model only has the heterogeneity of the mean; if no element in Θ is significantly different from zero, it means that the mean and square heterogeneous random parameter model degenerates into the general random parameter model.

[0103] S3: Calculate the leave-one-out cross-validation information criterion value for each of the random parameter models using the calculation module.

[0104] The leave-out cross-validation information criterion (LOOIC) value is twice the expected log prediction density of leave-out cross-validation. Specifically, in step S3, the calculation module obtains the following formula for the LOOIC value:

[0105] LOOIC = -2ELPD LOO ,

[0106]

[0107] In the formula,

[0108] p M (y i |θ) represents the M probability in the regression model;

[0109] p M (θ|y -i ) represents the posterior value of the parameter θ, where the observation value y i is excluded.

[0110] S4: Select a random parameter model with a lower LOOIC value to analyze the influencing factors of the economic loss of the fishing boat accident, and obtain the positive or negative influence of various explanatory variables on the economic loss of the fishing boat accident.

[0111] The following table shows the observed information of the sample characteristics of the fishing boat accident used in this embodiment. The results of the comparative analysis using these information are as follows:

[0112]

[0113]

[0114]

[0115] The estimated value is represented by the average value (standard deviation); the Bayesian confidence interval is [2.50%, 97.50%], and 0 is not included; ** indicates that the significance level of 5% is reached. As can be seen from the table, the LOOIC value of the RPMHM model is the smallest, and then the random parameter model is selected to analyze the influencing factors of the economic loss of the fishing boat accident, and the positive or negative influence of various explanatory variables on the economic loss of the fishing boat accident is obtained.

[0116] S5: Based on the positive or negative influence of various explanatory variables on the economic loss of the fishing boat accident obtained in step S4, the influence degree of the explanatory variables on the economic loss of the accident is excavated.

[0117] The accident economic loss influencing factor analysis method disclosed in this embodiment first constructs a Bayesian random parameter GB2 model through observed information. This model can solve the problem of highly non-normal distribution and heavy-tailed distribution. Then, based on this model, multiple random parameter models (RPM, RPMHM, and RPMHMV) are established, which can improve the statistical fitting degree and provide more insight. Then, the LOOIC value is used for effect comparison, the appropriate random parameter model is selected for analysis, and the positive or negative influence of various explanatory variables on the economic loss of the accident is obtained. Finally, based on the sensitivity analysis, the influence degree of the explanatory variables on the economic loss of the accident is excavated, thereby overcoming the problems of the existing analysis methods, such as the difficulty in handling the influencing factor analysis of the heavy-tailed distribution of the economic loss of the ship accident and the bias of the estimated parameters caused by the neglect of unobserved heterogeneity. The method is objective, comprehensive, and accurate, can excavate the influencing factors of the economic loss of the accident, has a wide application range, and better represents the influence of related factors on the economic loss caused by the accident.

[0118] Embodiment Two:

[0119] Referring to Figure 2 As shown in the figure, the embodiment further discloses an accident economic loss influencing factor analysis system based on embodiment one, which comprises a model module, a parameter model module, a calculation module, a selection module and an analysis module, wherein the parameter model module is electrically connected with the model module, the calculation module is electrically connected with the parameter model module, the selection module is electrically connected with the calculation module, and the analysis module is electrically connected with the selection module.

[0120] In the system, the model module is configured to construct a Bayesian random parameter GB2 model according to the observation information of the sample characteristics of the obtained fishing boat accidents; the parameter model module is configured to construct a plurality of random parameter models for analyzing the economic loss influencing factors of the fishing boat accidents based on the Bayesian random parameter GB2 model and the observation information of the sample characteristics; the calculation module is configured to calculate the LOOIC value of each random parameter model respectively; the selection module is configured to select the random parameter model with a lower LOOIC value to analyze the economic loss influencing factors of the fishing boat accidents, and obtain the positive or negative influence of various explanatory variables on the economic loss of the fishing boat accidents; and the analysis module is configured to analyze the positive or negative influence of various explanatory variables on the economic loss of the fishing boat accidents, and dig out the influence degree of the explanatory variables on the economic loss of the accidents.

[0121] The accident economic loss influencing factor analysis system disclosed in the embodiment solves the problems of highly non-normal distribution and heavy-tailed distribution by setting the model module to construct a Bayesian random parameter GB2 model based on the observation information; the parameter model module is configured to establish a plurality of random parameter models based on the model, which can improve the statistical fitting degree and provide more insights; the calculation module and the selection module are configured to use the leave-one-out cross-validation information criterion value for effect comparison, select the appropriate random parameter model for analysis, and obtain the positive or negative influence of various explanatory variables on the economic loss of the accidents; finally, the analysis module is based on sensitivity analysis to dig out the influence degree of the explanatory variables on the economic loss of the accidents, thereby overcoming the problems in the prior art, such as the difficulty in analyzing the influencing factors when the economic loss of the ship accident presents a heavy-tailed distribution, and the neglect of unobserved heterogeneity to cause bias in the estimated parameters, etc. The system is objective, comprehensive and accurate, can dig out the influencing factors of the accident economic loss, has a wide range of applications, and better represents the influence of related factors on the economic loss caused by the accident.

[0122] In the description of the embodiments of the present application, it should be noted that the terms "inner", "outer", and the like indicating the direction or positional relationship are based on the direction or positional relationship shown in the drawings, which is merely for the convenience of description, and does not indicate or imply that the device or member must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present application.

[0123] In the description of the present application, the description referring to the terms "one embodiment", "some embodiments", "in this embodiment", "specific example", or "some examples" and the like means that the specific features, mechanisms, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, mechanisms, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. Furthermore, different embodiments or examples described in the specification and the features of different embodiments or examples can be combined and combined by those skilled in the art without contradiction.

[0124] The above is merely a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An accident economic loss impact factor analysis method, characterized by, The method comprises the following steps: S1: constructing a Bayesian random parameter GB2 model according to observed information of sample characteristics of obtained fishing vessel accident; S2: constructing a plurality of random parameter models for analyzing economic loss influencing factors of the fishing vessel accident based on the Bayesian random parameter GB2 model and the observed information of the sample characteristics; S3: calculating a leave-one-out cross-validation information criterion value of each random parameter model by a calculation module; S4: selecting a random parameter model with a lower leave-one-out cross-validation information criterion value to analyze the economic loss influencing factors of the fishing vessel accident, and obtaining positive or negative influences of various explanatory variables on the economic loss of the fishing vessel accident; S5: based on the positive or negative influences of various explanatory variables on the economic loss of the fishing vessel accident obtained in the step S4, the influence degree of the explanatory variables on the economic loss of the accident is mined; In the Bayesian random parameter GB2 model, the probability density function of the random variable is as follows: In the formula, a is a shape parameter, which determines the peak value and tail thickness of the economic loss distribution; p is the probability of an accident with a small economic loss; q is the probability of an accident with a major economic loss; B(p, q) is a B function; λ reflects the economic loss level of the accident, which is a scale parameter, and controls the overall level and range of the loss, and the calculation formula is as follows: x is a covariate vector influencing the economic loss of the fishing vessel accident; β is a coefficient vector corresponding to x; k is the number of covariates; and the k-th moment of the random variable is as follows: The plurality of random parameter models in the step S2 comprise a general random parameter model, a mean heterogeneity random parameter model and a mean and variance heterogeneity random parameter model. The general random parameter model is expressed as follows: In the formula, b is a mean parameter vector of all observed values of the sample characteristics of the fishing vessel accident. The mean heterogeneity random parameter model is expressed as follows: In the formula, The mean and variance heterogeneity random parameter model is expressed as follows: λ = exp(β'x) = exp(β0+ β1x1+ β2x2+... + β k x k ), where In the formula, If there is no element in ω that is significantly different from zero, it means that the mean and variance heterogeneity random parameter model only has heterogeneity of the mean. If there is no element in Θ that is significantly different from zero, it means that the mean and variance heterogeneity random parameter model degenerates into the general random parameter model. In the step S3, the calculation formula of the leave-one-out cross-validation information criterion value obtained by the calculation module is as follows:

2. The method of claim 1, wherein, In the formula, 3. The method of claim 2, wherein, Based on the accident economic loss influencing factor analysis method in any one of claims 1-6, a circuit connection relationship is established: A model module is configured to construct a Bayesian random parameter GB2 model according to observed information of sample characteristics of obtained fishing vessel accidents; β n is a random parameter vector; A parameter model module is configured to construct a plurality of random parameter models for analyzing economic loss influencing factors of the fishing vessel accident based on the Bayesian random parameter GB2 model and the observed information of the sample characteristics; are random distributed error terms to capture possible heterogeneity.

4. The method of claim 3, wherein, A calculation module is configured to calculate a LOOIC value of each random parameter model; A selection module is configured to select a random parameter model with a lower leave-one-out cross-validation information criterion value to analyze the economic loss influencing factors of the fishing vessel accident, and obtain positive or negative influences of various explanatory variables on the economic loss of the fishing vessel accident; Z n To affect β n The explanatory variable vector of the mean; Θ is Z n a corresponding estimable parameter vector; if an element in Θ differs significantly from zero, then the mean-heterogeneity random parameter model is said to have mean heterogeneity.

5. The method of claim 4, wherein, ​ ​ W n is a vector of explanatory variables capturing the heterogeneity in the standard deviation σ; ω is W n corresponding estimable parameter vector; ​ ​ 6. The method of claim 5, wherein, ​ LOOIC = -2ELPD LOO , ​ p M (y i |θ) represents the M likelihood in the regression model; p M (θ|y -i ) represents the posterior value of the parameter θ, given the observation y i is excluded.

7. An accident economic loss impact factor analysis system characterized by, ​ ​ ​ ​ ​ The analysis module is configured to analyze the positive or negative influence of the various explanatory variables on the economic loss of the fishing vessel accident and to determine the influence degree of the explanatory variables on the economic loss of the accident.

Images