CN113298438A - Regional risk level assessment method and device, computer equipment and storage medium - Google Patents
Regional risk level assessment method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN113298438A CN113298438A CN202110694076.9A CN202110694076A CN113298438A CN 113298438 A CN113298438 A CN 113298438A CN 202110694076 A CN202110694076 A CN 202110694076A CN 113298438 A CN113298438 A CN 113298438A
- Authority
- CN
- China
- Prior art keywords
- risk
- risk rating
- index data
- submodel
- rating
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000012549 training Methods 0.000 claims abstract description 69
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 34
- 238000012502 risk assessment Methods 0.000 claims abstract description 34
- 238000011156 evaluation Methods 0.000 claims abstract description 32
- 238000012545 processing Methods 0.000 claims description 35
- 238000010606 normalization Methods 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 10
- 238000012937 correction Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 6
- 230000002265 prevention Effects 0.000 claims description 5
- 238000007477 logistic regression Methods 0.000 claims description 4
- 238000012706 support-vector machine Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 description 16
- 230000002349 favourable effect Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000013145 classification model Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 241000282412 Homo Species 0.000 description 1
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a method and a device for evaluating regional risk level, computer equipment and a storage medium, which can improve regional risk evaluation accuracy. The method comprises the following steps: initializing target area data to be subjected to area risk assessment to obtain index data of each preset dimension, wherein the index data comprise first index data under category characteristics and second index data under continuous characteristics; training based on a multi-modal heterogeneous algorithm to obtain a risk rating model meeting a preset training standard, wherein the risk rating model comprises a category characteristic risk rating sub-model used for identifying first index data and a continuous characteristic risk rating sub-model used for identifying second index data; respectively inputting the first index data and the second index data into a category characteristic risk rating submodel and a continuous characteristic risk rating submodel to obtain the prediction probability that the target area belongs to different preset risk levels; and determining the preset risk grade with the highest corresponding prediction probability as the risk evaluation grade of the target area.
Description
Technical Field
The present application relates to the field of artificial intelligence technologies, and in particular, to a method and an apparatus for evaluating a regional risk level, a computer device, and a storage medium.
Background
With the rapid development of artificial intelligence, the term "wisdom" is no longer used to describe humans, but also a city or a country. From government decision and service, to the living mode of people's clothing and eating and housing, to the industrial layout and planning of cities and to the operation and management mode of cities, all of which will move towards "intellectualization" under the support of artificial intelligence, the risk monitoring network of a smart city is an important construction for preventing city risks, the city will bear more and more people in the future, and the construction of the city risk monitoring network is important for ensuring city safety. The occurrence of a city risk accident is not an isolated event, but the result of a series of events occurring in succession. The environment is from natural environment, social environment, human defect, human unsafe behavior to object unsafe state, accident, injury, and the ring is buckled.
In the prior art, a complete and reliable risk assessment method for regional risks is not available, the risk assessment is abstract, and relevant departments are difficult to accurately grasp the application condition of the risk assessment, so that the regional risk level assessment is low in efficiency and accuracy.
Disclosure of Invention
In view of this, the present application provides a method and an apparatus for evaluating an area risk level, a computer device, and a storage medium, which can be used to solve the technical problems of low efficiency and low accuracy of area risk level evaluation when the area risk evaluation is performed at present.
According to one aspect of the present application, there is provided a method for assessing a regional risk level, the method comprising:
initializing target area data to be subjected to area risk assessment to obtain index data of each preset dimension, wherein the index data at least comprise first index data under category characteristics and second index data under continuous characteristics;
training based on a multi-modal heterogeneous algorithm to obtain a risk rating model meeting a preset training standard, wherein the risk rating model comprises a category characteristic risk rating sub-model suitable for the first index data identification and a continuous characteristic risk rating sub-model suitable for the second index data identification;
inputting the first index data and the second index data into the category characteristic risk rating submodel and the continuous characteristic risk rating submodel respectively to obtain the prediction probability that the target area belongs to different preset risk levels;
and determining the preset risk grade with the highest corresponding prediction probability as the risk evaluation grade of the target area.
According to another aspect of the present application, there is provided an apparatus for assessing a regional risk level, the apparatus including:
the processing module is used for initializing target area data to be subjected to area risk assessment to obtain index data of each preset dimension, wherein the index data at least comprise first index data under category characteristics and second index data under continuous characteristics;
the training module is used for training based on a multi-modal heterogeneous algorithm to obtain a risk rating model meeting a preset training standard, and the risk rating model comprises a category characteristic risk rating sub-model suitable for the first index data identification and a continuous characteristic risk rating sub-model suitable for the second index data identification;
the input module is used for respectively inputting the first index data and the second index data into the category characteristic risk rating submodel and the continuous characteristic risk rating submodel to obtain the prediction probability that the target area belongs to different preset risk levels;
and the determining module is used for determining a preset risk grade with the highest corresponding prediction probability as the risk evaluation grade of the target area.
According to yet another aspect of the present application, there is provided a non-transitory readable storage medium having stored thereon a computer program which, when executed by a processor, implements the above-described method of assessing a regional risk level.
According to yet another aspect of the present application, there is provided a computer device comprising a non-volatile readable storage medium, a processor, and a computer program stored on the non-volatile readable storage medium and executable on the processor, the processor implementing the above method for assessing a risk level of an area when executing the program.
By means of the technical scheme, compared with the existing regional risk level evaluation mode, the regional risk level evaluation method, the regional risk level evaluation device, the computer equipment and the storage medium can perform grid division on a target region to be subjected to risk evaluation, determine and obtain each grid unit, and extract index data of multiple preset dimensions in each grid unit. Based on a multi-modal heterogeneous algorithm capable of avoiding overfitting, a plurality of category characteristic risk rating submodels capable of being used for identifying category characteristics in index data and continuous characteristic risk rating submodels capable of identifying continuous characteristics in the index data are created and trained, different algorithms are randomly selected when different data batches are trained, and then overfitting of the models can be effectively prevented. For a new grid unit, index data in the grid unit can be divided into category features and continuous features, the category features and the continuous features are input into a plurality of models corresponding to a category feature risk rating submodel and a continuous feature risk rating submodel respectively, prediction probabilities of the grid unit belonging to different risk levels are obtained comprehensively, and finally the risk rating with the maximum prediction probability is used as the final risk evaluation level of the grid. In the scheme, risk assessment is carried out by using index data containing multiple dimensions such as risk factors, favorable factors, vulnerability and precautionary capacity, so that the risk of one area can be considered comprehensively, and the calculation result is more accurate. And through an analytic hierarchy process, weighting and summing are carried out on each risk factor, a scientific method is introduced to determine the weight, and in this respect, the calculation result is more accurate and reasonable. Therefore, by the technical scheme, the efficiency of regional risk assessment can be improved, and the assessment result of the regional risk level is more accurate.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application to the disclosed embodiment. In the drawings:
fig. 1 is a schematic flowchart illustrating an evaluation method for regional risk levels according to an embodiment of the present disclosure;
fig. 2 is a schematic flow chart illustrating another method for assessing a regional risk level according to an embodiment of the present application;
fig. 3 is a schematic structural diagram illustrating an apparatus for assessing a regional risk level according to an embodiment of the present application;
fig. 4 shows a schematic structural diagram of another device for assessing a regional risk level according to an embodiment of the present application.
Detailed Description
The present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Aiming at the technical problems of low efficiency and low accuracy of regional risk level evaluation caused by regional risk evaluation at present, the application provides a regional risk level evaluation method, as shown in fig. 1, the method comprises the following steps:
101. and initializing target area data to be subjected to area risk assessment to obtain index data of each preset dimension, wherein the index data at least comprises first index data under category characteristics and second index data under continuous characteristics.
The target area data is data extracted from a target area to be subjected to area risk assessment, and when the target area data is extracted, a plurality of preset dimensions influencing area risks can be determined at first, and specifically, data acquisition can be performed from a plurality of dimensions such as risk factors, favorable factors, and precautionary capacity (such as area GDP, supporting facilities, medical level). In the scheme, the following index data of five dimensions can be obtained: (1) and natural disaster data: seismic data, typhoon data, debris flow, flood and other data; (2) and human data: economy (GDP data), population (population density); (3) and high risk factor: quantitative data of chemical plants, hydropower stations, fireworks plants and the like; (4) favorable factors: quantity data of fire stations, hospitals, ambulance stations and the like; (5) and insurance data: target address, premium data, pay data; wherein the natural disaster data comprises the frequency and intensity of the historical disasters in the area; in order to ensure the reference value of the preset dimension data, the insurance data only needs to count the data of the last 2 years.
For this embodiment, in order to ensure the influence of each index data dimension, the initialization processing may specifically include normalization processing, and the index data in each dimension is controlled in a preset value interval through the normalization processing, so as to facilitate analysis processing of the index data. After normalization processing, longitude and latitude coordinate correction can be performed on index data (such as an address of an insurance target, an address of a natural disaster target in natural disaster data, and the like) carrying position information. Correspondingly, the index data has different influences on the classification and division due to different corresponding data forms, and specifically may include discrete data (e.g., discrete data such as presence, absence, and quantity of high risk factors and beneficial factors) and continuous data (e.g., continuous numerical values such as GDP data, population density data, premium data, and pay data). Therefore, in the present application, after the normalization processing and the correction processing of the longitude and latitude coordinates are performed on the index data, classification of category features and continuous features is performed on the index data, and the index data is further classified into first index data under the category features and second index data under the continuous features.
102. And training based on a multi-modal heterogeneous algorithm to obtain a risk rating model meeting a preset training standard, wherein the risk rating model comprises a category characteristic risk rating sub-model suitable for first index data identification and a continuous characteristic risk rating sub-model suitable for second index data identification.
For the embodiment, a risk rating model may be first constructed and trained, and in order to facilitate accurate identification of different index data, the risk rating model may include a category feature risk rating sub-model suitable for identification of first index data and a continuous feature risk rating sub-model suitable for identification of second index data. In a specific application scenario, the sample label data can be used for respectively training a category characteristic risk rating submodel and a continuous characteristic risk rating submodel, and then the trained category characteristic risk rating submodel and the trained continuous characteristic risk rating submodel are comprehensively constructed into a risk rating model, so that the risk rating model is used for realizing the risk precision evaluation of target area data under different indexes.
103. And respectively inputting the first index data and the second index data into the category characteristic risk rating submodel and the continuous characteristic risk rating submodel to obtain the prediction probability that the target area belongs to different preset risk levels.
The preset risk level may be specifically divided according to an actual application scenario, for example, the preset risk level may be divided into three evaluation levels of high risk, medium risk, and low risk according to the severity, and further, more detailed division may be performed, for example, the preset risk level may be divided into five levels of high risk, medium risk, low risk, and low risk. According to the evaluation level setting of the preset risk level, the category characteristic risk rating submodel and the continuous characteristic risk rating submodel correspond to different output results, when the preset risk level corresponds to three evaluation levels, the output result of the risk rating model is a vector with the length of 3, and the output result respectively represents the prediction probability of belonging to 3 different preset risk levels; when the preset risk level corresponds to five evaluation levels, the output result of the risk rating model is a vector with the length of 5, and the vector represents the prediction probability of 5 different preset risk levels. In the present application, the preset risk level is preferably divided into five levels of high risk, higher risk, medium risk, lower risk and low risk.
In a specific application scenario, after it is determined that the risk rating model is trained, the index data can be further input into the risk rating model according to the index types, that is, the first index data and the second index data are respectively input into the category characteristic risk rating submodel and the continuous characteristic risk rating submodel to respectively obtain prediction results under the corresponding index data, and then the prediction results of the category characteristic and the continuous characteristic are summed up to determine the prediction probability that the target area belongs to different preset risk levels.
For example, if the index data in the target area includes a, b, c, d, e, and f, the index data is further divided into first index data a, b, c, and d under the category characteristic and second index data e and f under the continuous characteristic according to the index type. When the risk level of the target area is evaluated, the first index data a, b, c and d can be input into the category characteristic risk rating submodel to obtain first prediction probabilities that the target area belongs to different preset risk levels, the second index data e and f are input into the continuous characteristic risk rating submodel to obtain second prediction probabilities that the target area belongs to different preset risk levels, and then the first prediction probability and the second prediction probability can be added for calculation, so that the final prediction probabilities that the target area belongs to different preset risk levels can be calculated.
104. And determining the preset risk grade with the highest corresponding prediction probability as the risk evaluation grade of the target area.
For this embodiment, if the prediction probabilities of the target region belonging to different preset risk levels obtained based on the risk assessment model are respectively: the high risk is 40%, the higher risk is 10%, the middle risk is 20%, the lower risk is 15% and the low risk is 15%, through comparative analysis, the prediction probability corresponding to the high risk can be determined to be the highest, and therefore the high risk can be determined as the risk assessment grade of the target area.
By the method for evaluating the regional risk level in this embodiment, each grid unit can be determined and obtained by performing grid division on a target region to be subjected to risk evaluation, and index data of a plurality of preset dimensions in each grid unit can be extracted. Based on a multi-modal heterogeneous algorithm capable of avoiding overfitting, a plurality of category characteristic risk rating submodels capable of being used for identifying category characteristics in index data and continuous characteristic risk rating submodels capable of identifying continuous characteristics in the index data are created and trained, different algorithms are randomly selected when different data batches are trained, and then overfitting of the models can be effectively prevented. For a new grid unit, index data in the grid unit can be divided into category features and continuous features, the category features and the continuous features are input into a plurality of models corresponding to a category feature risk rating submodel and a continuous feature risk rating submodel respectively, prediction probabilities of the grid unit belonging to different risk levels are obtained comprehensively, and finally the risk rating with the maximum prediction probability is used as the final risk evaluation level of the grid. In the scheme, risk assessment is carried out by using index data containing multiple dimensions such as risk factors, favorable factors, vulnerability and precautionary capacity, so that the risk of one area can be considered comprehensively, and the calculation result is more accurate. And through an analytic hierarchy process, weighting and summing are carried out on each risk factor, a scientific method is introduced to determine the weight, and in this respect, the calculation result is more accurate and reasonable. Therefore, by the technical scheme, the efficiency of regional risk assessment can be improved, and the assessment result of the regional risk level is more accurate.
Further, as a refinement and an extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process in this embodiment, another method for evaluating a regional risk level is provided, as shown in fig. 2, and the method includes:
201. and performing grid division processing on target area data in the target area to obtain a plurality of grid units with preset sizes, wherein the grid units contain grid data with each preset dimension.
For this embodiment, after the target area is determined, the target area may be first subjected to area rasterization, and the target area is divided into a plurality of grid units with preset sizes according to a certain rule, so as to determine the risk level of each grid unit, for example, it may be determined that the risk of the area is five levels, such as high risk, medium risk, low risk, and low risk. Common grid division includes area division of roads and water systems or standard rectangle division according to longitude and latitude. The proposed solution uses longitude and latitude to rasterize the region, where the resolution is 2 km, i.e. the target region is divided into 2 x 2 squares.
After each grid unit is determined, grid data in each grid unit can be further extracted, and specifically, data acquisition can be performed from multiple dimensions such as risk factors, benefit factors, and precautionary capabilities (such as regional GDP, supporting facilities, medical levels). In the scheme, the following index data of five dimensions can be obtained: (1) and natural disaster data: seismic data, typhoon data, debris flow, flood and other data; (2) and human data: economy (GDP data), population (population density); (3) and high risk factor: quantitative data of chemical plants, hydropower stations, fireworks plants and the like; (4) favorable factors: quantity data of fire stations, hospitals, ambulance stations and the like; (5) and insurance data: target address, premium data, pay data; wherein the natural disaster data comprises the frequency and intensity of the historical disasters in the area; in order to ensure the reference value of the preset dimension data, the insurance data only needs to count the data of the last 2 years.
202. And performing initialization processing on each grid data to obtain first index data related to the category characteristics and second index data related to the continuous characteristics in each grid unit, wherein the initialization processing at least comprises normalization processing, coordinate correction processing and classification processing of the category characteristics and the continuous characteristics.
For the embodiment, after obtaining the grid data, a series of processes of normalization, longitude and latitude coordinate correction, and classification of the category feature and the continuous feature may be performed on the grid data. When normalization processing is performed, a min-max standardized processing mode can be adopted, and the specific form is as follows:
wherein Xmin、XmaxRespectively representing the maximum value and the minimum value corresponding to the index data X.
After normalization processing, longitude and latitude coordinate correction can be performed on index data (such as an address of an insurance target, an address of a natural disaster target in natural disaster data, and the like) carrying position information. In a specific application scenario, in order to assign a target address to a corresponding grid cell, latitude and longitude information of the target address needs to be acquired. In a specific application scenario, the longitude and latitude information of the address can be specifically analyzed by adopting the existing map navigation application. For the same address, different map navigation applications can be adopted to respectively analyze the target address, and then longitude and latitude information with finer geographic levels is selected and analyzed in a contrast mode to serve as a final longitude and latitude coordinate.
Correspondingly, the raster data has different influences on classification due to different corresponding data forms, and specifically may include discrete data (e.g., discrete data such as presence, absence, and quantity of high risk factors and beneficial factors) and continuous data (e.g., continuous numerical values such as GDP data, population density data, premium data, and pay data). In order to ensure the accuracy of model identification, in the application, after normalization and longitude and latitude coordinate correction processing are performed on the raster data, classification of category features and continuous features is also required to be performed on the raster data, and further classification is performed to obtain first index data under the category features and second index data under the continuous features.
203. Respectively training a first risk rating sub-model and a second risk rating sub-model which are suitable for category feature recognition by using a CART classification tree algorithm and a naive Bayes algorithm, and respectively training a third risk rating sub-model and a fourth risk rating sub-model which are suitable for continuous feature recognition by using a multinomial logistic regression model and a support vector machine algorithm.
For the embodiment, a final prediction result can be determined and obtained based on a multi-modal heterogeneous algorithm, namely, a plurality of selectable and recognizable risk rating submodels are respectively trained for each index type, and the final prediction result is determined and obtained through integration of the estimation results of the plurality of risk rating submodels. In the application, aiming at the category characteristics, a first risk rating submodel and a second risk rating submodel can be respectively trained on the basis of a CART classification tree algorithm and a naive Bayes algorithm, namely the category characteristic risk rating submodel comprises the first risk rating submodel and the second risk rating submodel; for the continuous features, a third risk rating sub-model and a fourth risk rating sub-model may be trained based on the multiple logistic regression model and the support vector machine algorithm, respectively, even though the continuous feature risk rating sub-model includes the third risk rating sub-model and the fourth risk rating sub-model. In this embodiment, a plurality of risk rating submodels are set, and then the output results of the plurality of risk rating submodels are integrated to obtain a final risk evaluation result, so that the accuracy of risk prediction can be ensured. It should be noted that, in the present application, two risk rating submodels may be trained for category features and continuous features respectively based on a multi-modal heterogeneous algorithm, and it is easy for a person skilled in the art to think that one or more (more than two) risk rating submodels may also be trained for each feature type at the same time, thereby achieving the technical effect of the present application or being superior to the present application. Therefore, the technical scheme that one or more risk rating sub-models are trained by using other related machine learning algorithms for each feature type so as to ensure the risk assessment accuracy of the models is within the protection scope of the application and is not specifically limited herein.
204. And training the risk rating model in batches based on the sample label data and the first risk rating submodel, the second risk rating submodel, the third risk rating submodel and the fourth risk rating submodel so as to enable the risk rating model to accord with the preset training standard.
Wherein, the sample label data is marked with risk rating probability under each preset risk level.
In a specific application scenario, since the insurance claim data can more visually display the risk level result, in this embodiment, after the sample data is determined, the insurance claim rate can be calculated according to the historical insurance data in the sample data, that is:
in general VClaims payment<VPremium feeThus, R has a value in the range of [0,1 ]]If V isClaims payment＞VPremium feeAnd then, R is 1.
Further, the corresponding risk level can be determined according to the insurance claim rate, and specifically, the interval [0,1 ] can be determined]Divided into 5 cells [0, 0.2 ]]、[0.2，0.4]、[0.4，0.6]、[0.6，0.8]And [0.8, 1.0]And dividing the grid units into low risk, medium risk, high risk and high risk levels according to the located intervals of the claim ratio R. Aiming at 10000 randomly selected grid units, a data sample with a label can be obtained
Wherein Xi＝{xi,1,xi,2,…,xi,mDenotes each index data on the grid i; y isiRepresenting the risk level in the form of One-hot format.
Based on these tag data
The application provides a multi-mode heterogeneous algorithm for classification, and the main idea of the multi-mode heterogeneous algorithm is to process category features and continuous features respectively. In addition, to prevent algorithm overfitting and robustness of the stronger model, the input continuous features are perturbed randomly to enhance the data. The data enhancement mode is as follows: adding tiny random white noise to each characteristic in the continuous characteristics X _ cont, namely randomly adding normal distribution N (0,0.01) to each characteristic, expanding 5000 pieces of data and then combining the data with the original data to obtain sample label data after data enhancement
Further, the data-enhanced sample label data may be utilized to train a first risk rating submodel (denoted as F1), a second risk rating submodel (denoted as F2) adapted to the category features, and a third risk rating submodel (denoted as F3) and a fourth risk rating submodel (denoted as F4) adapted to the continuous features.
Specifically, since the data amount is large, the data batch training method can be adopted to divide the 15000 sample label data into training batches, for example, 15 batches, each batch containing 100 sample label data. And randomly selecting a risk rating sub-model for the category characteristics and the continuous characteristics in each batch for model training. During model training, the labeled data after data enhancement in each training batch can be divided into a training set and a verification set according to the proportion of 8:2, and the input features of the training set are divided into category features and continuous features. Randomly selecting one category characteristic risk rating submodel F _ category as a category characteristic risk rating submodel for training and identifying the category characteristic X _ category under the current data batch from the first risk rating submodel and the second risk rating submodel, and randomly selecting one continuous characteristic risk rating submodel F _ cont as a continuous characteristic risk rating submodel for training and identifying the continuous characteristic X _ cont under the current data batch from the third risk rating submodel and the fourth risk rating submodel to obtain:
F＝w1*F_category(X_category)+w2*F_cont(X_cont)
where w1 and w2 represent the weights of F _ category and F _ cont, respectively, setting w1: w2 as the number of class features: a number of consecutive features; f is a length 5 vector representing the probability of belonging to five different preset risk levels, respectively.
Since the target risk level data is usually unbalanced (the number of grids at high risk level is generally much less), in order to improve the accuracy of the model, the Focal local function is used as the Loss function of the model, and the Focal local function is used as the objective function, that is, the Loss function is:
wherein, Fi,jRepresenting model prediction ith sample belonging to jth categoryProbability of yi,jRepresenting the true probability that the ith sample belongs to the jth sample; the parameters alpha and gamma take values of 0.25 and 2 respectively, and the following minimization problem is solved by adopting an adam optimization solver:
to determine a parameter θ;
for the embodiment, in the risk rating model training process, when it is determined that the loss functions of the risk rating submodels in each training batch reach the convergence state, or it is determined that the numerical values of the loss functions are all smaller than the preset threshold, it may be determined that the first risk rating submodel, the second risk rating submodel, the third risk rating submodel, and the fourth risk rating submodel conform to the preset training standard, and further, the trained first risk rating submodel, the trained second risk rating submodel, the trained third risk rating submodel, and the trained fourth risk rating submodel may be combined to construct a risk rating model. It should be noted that, because the data volume of the tagged data after data enhancement is large, during training, the tagged data may be divided into data batches, and for each data batch, different category feature risk rating submodels F _ category and continuous feature risk rating submodels F _ cont are respectively selected at random, so that the overfitting problem of the model can be effectively prevented.
Correspondingly, the embodiment step 204 may specifically include: dividing sample label data into a plurality of training batches; for each training batch, determining any one of the first risk rating submodel and the second risk rating submodel as a category characteristic risk rating submodel, and determining any one of the third risk rating submodel and the fourth risk rating submodel as a continuous characteristic risk rating submodel; training a category characteristic risk rating submodel and a continuous characteristic risk rating submodel under each training batch according to the risk rating probability marked in the sample label data; and if the loss functions of the category characteristic risk rating submodel and the continuous characteristic risk rating submodel are smaller than a preset threshold value under any training batch, judging that the risk rating model completes training.
205. And respectively inputting the first index data into the first risk rating submodel and the second risk rating submodel to obtain a first prediction result and a second prediction result.
For this embodiment, after the first index data of the target region for the category features is extracted, the first index data may be simultaneously input into the first risk rating submodel and the second risk rating submodel, and different prediction results of the two risk rating submodels for the first index data are obtained.
206. And inputting the second index data into a third risk rating submodel and a fourth risk rating submodel respectively to obtain a third prediction result and a fourth prediction result.
For this embodiment, after the second index data of the target region for the continuous features is extracted, the second index data may be simultaneously input into the third risk rating submodel and the fourth risk rating submodel, and different prediction results of the two risk rating submodels for the second index data are obtained.
207. And calculating the prediction probability of each grid unit aiming at each preset risk level according to the first prediction result, the second prediction result, the third prediction result and the fourth prediction result.
For the embodiment, after the first prediction result and the second prediction result for the first index data, and the third prediction result and the fourth prediction result for the second index data are obtained through calculation, an integrated result of the first prediction result, the second prediction result, the third prediction result and the fourth prediction result may be obtained through calculation according to an integrated calculation formula, and the integrated result is further determined as the prediction probability of the grid unit for each preset risk level.
The integrated calculation formula is characterized by:
P＝0.5(F1(X_category)+F2(X_category)+F3(X_cont)+F4(X_cont))
wherein P is a vector with a length of 5, and since F1 and F2 both belong to the prediction results under the category features, and F3 and F4 both belong to the prediction results under the continuous features, when the probabilities of grid cells belonging to different risk levels are calculated comprehensively, the sum of the probabilities of the first risk rating submodel, the second risk rating submodel, the third risk rating submodel and the fourth risk rating submodel needs to be entirely divided by 2, that is, after the sum of the four risk submodels is obtained by calculation, the sum is multiplied by 0.5, so that the probabilities of the grid cells belonging to five different risk levels (high, medium, low and low) can be obtained, and for this embodiment, the risk rating with the highest probability can be used as the final risk level of the grid cell.
Correspondingly, as an optional mode, the configuration weights of the first risk rating submodel, the second risk rating submodel, the third risk rating submodel and the fourth risk rating submodel in the integrated calculation formula are all 50%. However, in a specific application scenario, different weights may be configured for different risk rating submodels according to the prediction accuracy of each model. For example, for a first risk rating submodel, the configurable weight is a, the configurable weight of a second risk rating submodel is (1-a), for a third risk rating submodel, the configurable weight is b, and the configurable weight of a fourth risk rating submodel is (1-b), where a and b are both values greater than 0 and less than or equal to 1.
Accordingly, the characterization of the integrated calculation formula described above can be transformed into:
P＝a*F1(X_category)+(1-a)*F2(X_category)+b*F3(X_cont)+(1-b)*F4(X_cont)
208. and determining the preset risk level with the highest corresponding prediction probability as the risk evaluation level of each grid unit.
In a specific application scenario, as an optimal mode, after determining and obtaining the risk assessment level of each grid unit in a target area, risk marking can be further performed on the grid units according to the risk assessment level of each grid unit in the target area; and generating a risk prevention and control recommendation aiming at the grid unit with the risk evaluation grade larger than a preset risk threshold value.
Specifically, after determining the risk level corresponding to each grid unit in the target region, a risk distribution map may be generated according to the risk level, and color marking may be performed on each grid region in the risk distribution map according to the risk level, for example, red, orange, yellow, blue, and green may be used as the early warning colors corresponding to the high, medium, low, and low five-level early warning levels, and color grading may be performed by using a natural breakpoint method, a standard deviation method, an equidistant segmentation method, and the like, which is not limited herein. The risk distribution situation can be more intuitively displayed through the risk distribution map, and early warning analysis on the risk is facilitated. So that the prediction of the risk level of the area is made before no risk occurs in the target area, so that prevention and maintenance can be performed in advance. Further, for an area with a higher risk level, a corresponding risk prevention and control strategy can be generated, for example, a target insurance product can be matched according to the feature information corresponding to the index data of the five dimensions, and then a target insurance product recommendation can be generated for the target area.
By means of the regional risk level assessment method, each grid unit can be determined and obtained by performing grid division on a target region to be subjected to risk assessment, and index data of a plurality of preset dimensions in each grid unit are extracted. Based on a multi-modal heterogeneous algorithm capable of avoiding overfitting, a category characteristic risk rating sub-model capable of being used for identifying category characteristics in index data and a continuous characteristic risk rating sub-model capable of identifying continuous characteristics in the index data are created and trained, different algorithms are randomly selected when different data batches are trained, and then the overfitting problem of the models can be effectively prevented. For a new grid unit, index data in the grid unit can be divided into class features and continuous features, the class features and the continuous features are input into a plurality of models corresponding to a class feature classification model and a continuous feature classification model respectively, prediction probabilities of the grid unit belonging to different risk levels are obtained through synthesis, and finally a risk grade with the maximum prediction probability is used as a final risk evaluation grade of the grid. In the scheme, risk assessment is carried out by using index data containing multiple dimensions such as risk factors, favorable factors, vulnerability and precautionary capacity, so that the risk of one area can be considered comprehensively, and the calculation result is more accurate. And through an analytic hierarchy process, weighting and summing are carried out on each risk factor, a scientific method is introduced to determine the weight, and in this respect, the calculation result is more accurate and reasonable. Therefore, by the technical scheme, the efficiency of regional risk assessment can be improved, and the assessment result of the regional risk level is more accurate.
Further, as a specific implementation of the method shown in fig. 1 and fig. 2, an embodiment of the present application provides an apparatus for assessing a regional risk level, as shown in fig. 3, the apparatus includes: a processing module 31, a training module 32, an input module 33, and a determination module 34;
the processing module 31 is configured to perform initialization processing on target area data to be subjected to area risk assessment to obtain index data of each preset dimension, where the index data at least includes first index data under category characteristics and second index data under continuous characteristics;
the training module 32 is used for training based on a multi-modal heterogeneous algorithm to obtain a risk rating model meeting a preset training standard, wherein the risk rating model comprises a category characteristic risk rating sub-model suitable for first index data identification and a continuous characteristic risk rating sub-model suitable for second index data identification;
the input module 33 is configured to input the first index data and the second index data into the category characteristic risk rating submodel and the continuous characteristic risk rating submodel, respectively, to obtain prediction probabilities that the target area belongs to different preset risk levels;
the determining module 34 may be configured to determine a preset risk level with the highest corresponding prediction probability as the risk assessment level of the target area.
In a specific application scenario, in order to obtain index data of each preset dimension based on target area data, the processing module 31 may be specifically configured to perform grid division processing on the target area data in the target area to obtain a plurality of grid units of a preset size, where each grid unit includes grid data of each preset dimension; and performing initialization processing on each grid data to obtain first index data related to the category characteristics and second index data related to the continuous characteristics in each grid unit, wherein the initialization processing at least comprises normalization processing, coordinate correction processing and classification processing of the category characteristics and the continuous characteristics.
Correspondingly, when the risk rating model meeting the preset training standard is trained, the training module 32 may be specifically configured to respectively train a first risk rating sub-model and a second risk rating sub-model suitable for category feature recognition by using the CART classification tree algorithm and the naive bayes algorithm; respectively training a third risk rating sub-model and a fourth risk rating sub-model which are suitable for continuous feature recognition by utilizing a multinomial logistic regression model and a support vector machine algorithm; and training the risk rating model in batches based on the sample label data and the first risk rating submodel, the second risk rating submodel, the third risk rating submodel and the fourth risk rating submodel so as to enable the risk rating model to accord with the preset training standard.
In a specific application scenario, the sample label data is marked with risk rating probabilities at each preset risk level; correspondingly, the training module 32 may be specifically configured to divide the sample label data into a plurality of training batches when the risk rating model is trained in batches based on the sample label data and the first risk rating submodel, the second risk rating submodel, the third risk rating submodel, and the fourth risk rating submodel so that the risk rating model meets a preset training standard; for each training batch, determining any one of the first risk rating submodel and the second risk rating submodel as a category characteristic risk rating submodel, and determining any one of the third risk rating submodel and the fourth risk rating submodel as a continuous characteristic risk rating submodel; training a category characteristic risk rating submodel and a continuous characteristic risk rating submodel under each training batch according to the risk rating probability marked in the sample label data; and if the loss functions of the category characteristic risk rating submodel and the continuous characteristic risk rating submodel are smaller than a preset threshold value under any training batch, judging that the risk rating model completes training.
Correspondingly, in order to determine and obtain the prediction probabilities that the target area belongs to different preset risk levels, the input module 33 may be specifically configured to input the first index data into the first risk rating submodel and the second risk rating submodel, respectively, and obtain a first prediction result and a second prediction result; inputting the second index data into a third risk rating submodel and a fourth risk rating submodel respectively to obtain a third prediction result and a fourth prediction result; and calculating the prediction probability of each grid unit aiming at each preset risk level according to the first prediction result, the second prediction result, the third prediction result and the fourth prediction result.
In a specific application scenario, in order to determine the risk assessment level of the target region based on the prediction probabilities that the target region belongs to different preset risk levels, the determining module 34 may be specifically configured to determine a preset risk level with the highest corresponding prediction probability as the risk assessment level of each grid unit.
Accordingly, in order to evaluate the risk level based on the risk of each grid unit in the target area, as shown in fig. 4, the apparatus further includes: a marking module 35, a generating module 36;
the marking module 35 is configured to mark risks of the grid units according to the risk assessment levels of the grid units in the target area;
and the generating module 36 may be configured to generate a risk prevention and control recommendation for the grid unit with the risk assessment level greater than the preset risk threshold.
It should be noted that other corresponding descriptions of the functional units related to the apparatus for evaluating a regional risk level provided in this embodiment may refer to the corresponding descriptions in fig. 1 to fig. 2, and are not repeated herein.
Based on the methods shown in fig. 1 to 2, correspondingly, the present embodiment further provides a non-volatile storage medium, on which computer readable instructions are stored, and the computer readable instructions, when executed by a processor, implement the method for assessing the risk level of the area shown in fig. 1 to 2.
Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.), and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method of the embodiments of the present application.
Based on the method shown in fig. 1 to fig. 2 and the virtual device embodiments shown in fig. 3 and fig. 4, in order to achieve the above object, the present embodiment further provides a computer device, where the computer device includes a storage medium and a processor; a nonvolatile storage medium for storing a computer program; a processor for executing a computer program to implement the above-described method for assessing a regional risk level as shown in fig. 1-2.
Optionally, the computer device may further include a user interface, a network interface, a camera, Radio Frequency (RF) circuitry, a sensor, audio circuitry, a WI-FI module, and so forth. The user interface may include a Display screen (Display), an input unit such as a keypad (Keyboard), etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface), etc.
It will be understood by those skilled in the art that the present embodiment provides a computer device structure that is not limited to the physical device, and may include more or less components, or some components in combination, or a different arrangement of components.
The nonvolatile storage medium can also comprise an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the computer device described above, supporting the operation of information handling programs and other software and/or programs. The network communication module is used for realizing communication among components in the nonvolatile storage medium and communication with other hardware and software in the information processing entity device.
Through the above description of the embodiments, those skilled in the art will clearly understand that the present application can be implemented by software plus a necessary general hardware platform, and can also be implemented by hardware.
Through the technical scheme, compared with the prior art, the method and the device have the advantages that the target area to be subjected to risk assessment is subjected to grid division, each grid unit is determined to be obtained, and index data of a plurality of preset dimensions in each grid unit are extracted. Based on a multi-modal heterogeneous algorithm capable of avoiding overfitting, a plurality of category characteristic risk rating submodels capable of being used for identifying category characteristics in index data and continuous characteristic risk rating submodels capable of identifying continuous characteristics in the index data are created and trained, different algorithms are randomly selected when different data batches are trained, and then overfitting of the models can be effectively prevented. For a new grid unit, index data in the grid unit can be divided into category features and continuous features, the category features and the continuous features are input into a plurality of models corresponding to a category feature risk rating submodel and a continuous feature risk rating submodel respectively, prediction probabilities of the grid unit belonging to different risk levels are obtained comprehensively, and finally the risk rating with the maximum prediction probability is used as the final risk evaluation level of the grid. In the scheme, risk assessment is carried out by using index data containing multiple dimensions such as risk factors, favorable factors, vulnerability and precautionary capacity, so that the risk of one area can be considered comprehensively, and the calculation result is more accurate. And through an analytic hierarchy process, weighting and summing are carried out on each risk factor, a scientific method is introduced to determine the weight, and in this respect, the calculation result is more accurate and reasonable. Therefore, by the technical scheme, the efficiency of regional risk assessment can be improved, and the assessment result of the regional risk level is more accurate.
Those skilled in the art will appreciate that the figures are merely schematic representations of one preferred implementation scenario and that the blocks or flow diagrams in the figures are not necessarily required to practice the present application. Those skilled in the art will appreciate that the modules in the devices in the implementation scenario may be distributed in the devices in the implementation scenario according to the description of the implementation scenario, or may be located in one or more devices different from the present implementation scenario with corresponding changes. The modules of the implementation scenario may be combined into one module, or may be further split into a plurality of sub-modules.
The above application serial numbers are for description purposes only and do not represent the superiority or inferiority of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of the present application, but the present application is not limited thereto, and any variations that can be made by those skilled in the art are intended to fall within the scope of the present application.
Claims (10)
1. A method for assessing a regional risk level, comprising:
initializing target area data to be subjected to area risk assessment to obtain index data of each preset dimension, wherein the index data at least comprise first index data under category characteristics and second index data under continuous characteristics;
training based on a multi-modal heterogeneous algorithm to obtain a risk rating model meeting a preset training standard, wherein the risk rating model comprises a category characteristic risk rating sub-model suitable for the first index data identification and a continuous characteristic risk rating sub-model suitable for the second index data identification;
inputting the first index data and the second index data into the category characteristic risk rating submodel and the continuous characteristic risk rating submodel respectively to obtain the prediction probability that the target area belongs to different preset risk levels;
and determining the preset risk grade with the highest corresponding prediction probability as the risk evaluation grade of the target area.
2. The method according to claim 1, wherein the target area data to be subjected to area risk assessment is initialized to obtain index data of each preset dimension, the index data at least includes first index data under a category characteristic and second index data under a continuous characteristic, and specifically includes:
performing grid division processing on target area data in a target area to obtain a plurality of grid units with preset sizes, wherein the grid units contain grid data with each preset dimension;
and performing initialization processing on each grid data to obtain first index data related to the class characteristics and second index data related to the continuous characteristics in each grid unit, wherein the initialization processing at least comprises normalization processing, coordinate correction processing and classification processing of the class characteristics and the continuous characteristics.
3. The method according to claim 1, wherein the training based on the multi-modal heterogeneous algorithm obtains a risk rating model meeting a preset training standard, and the risk rating model comprises a category feature risk rating submodel suitable for the first index data recognition and a continuous feature risk rating submodel suitable for the second index data recognition, and specifically comprises:
respectively training a first risk rating sub-model and a second risk rating sub-model which are suitable for class feature recognition by utilizing a CART classification tree algorithm and a naive Bayes algorithm;
respectively training a third risk rating sub-model and a fourth risk rating sub-model which are suitable for continuous feature recognition by utilizing a multinomial logistic regression model and a support vector machine algorithm;
training a risk rating model in batches based on sample label data and the first, second, third and fourth risk rating submodels, so that the risk rating model meets a preset training standard.
4. The method of claim 3, wherein the sample label data is labeled with a risk rating probability at each preset risk level;
the training of the risk rating model in batches based on the sample tag data and the first, second, third and fourth risk rating submodels to make the risk rating model meet a preset training standard specifically comprises:
dividing the sample label data into a plurality of training batches;
for each training batch, determining any one of the first risk rating submodel and the second risk rating submodel as a category feature risk rating submodel, and determining any one of the third risk rating submodel and the fourth risk rating submodel as a continuous feature risk rating submodel;
training a category characteristic risk rating submodel and a continuous characteristic risk rating submodel under each training batch according to the risk rating probability marked in the sample label data;
and if the loss functions of the category characteristic risk rating submodel and the continuous characteristic risk rating submodel are smaller than a preset threshold value under any training batch, judging that the risk rating model completes training.
5. The method according to claim 4, wherein the inputting the first index data and the second index data into the category feature risk rating submodel and the continuous feature risk rating submodel respectively to obtain the predicted probabilities that the target region belongs to different preset risk levels specifically comprises:
inputting the first index data into the first risk rating submodel and the second risk rating submodel respectively to obtain a first prediction result and a second prediction result;
inputting the second index data into the third risk rating submodel and the fourth risk rating submodel respectively to obtain a third prediction result and a fourth prediction result;
and calculating the prediction probability of each grid unit aiming at each preset risk level according to the first prediction result, the second prediction result, the third prediction result and the fourth prediction result.
6. The method according to claim 5, wherein the determining the preset risk level corresponding to the highest predicted probability as the risk assessment level of the target area specifically comprises:
and determining the preset risk level with the highest corresponding prediction probability as the risk evaluation level of each grid unit.
7. The method of claim 1, further comprising:
according to the risk assessment level of each grid unit in the target area, carrying out risk marking on the grid unit;
and generating a risk prevention and control recommendation aiming at the grid unit with the risk evaluation grade larger than a preset risk threshold value.
8. An apparatus for assessing a risk level of a region, comprising:
the processing module is used for initializing target area data to be subjected to area risk assessment to obtain index data of each preset dimension, wherein the index data at least comprise first index data under category characteristics and second index data under continuous characteristics;
the training module is used for training based on a multi-modal heterogeneous algorithm to obtain a risk rating model meeting a preset training standard, and the risk rating model comprises a category characteristic risk rating sub-model suitable for the first index data identification and a continuous characteristic risk rating sub-model suitable for the second index data identification;
the input module is used for respectively inputting the first index data and the second index data into the category characteristic risk rating submodel and the continuous characteristic risk rating submodel to obtain the prediction probability that the target area belongs to different preset risk levels;
and the determining module is used for determining a preset risk grade with the highest corresponding prediction probability as the risk evaluation grade of the target area.
9. A non-transitory readable storage medium having stored thereon a computer program, wherein the program, when executed by a processor, implements the method of assessing a regional risk level of any of claims 1 to 7.
10. A computer device comprising a non-volatile readable storage medium, a processor and a computer program stored on the non-volatile readable storage medium and executable on the processor, wherein the processor implements the method of assessing a regional risk level of any of claims 1 to 7 when executing the program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110694076.9A CN113298438A (en) | 2021-06-22 | 2021-06-22 | Regional risk level assessment method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110694076.9A CN113298438A (en) | 2021-06-22 | 2021-06-22 | Regional risk level assessment method and device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113298438A true CN113298438A (en) | 2021-08-24 |
Family
ID=77329125
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110694076.9A Pending CN113298438A (en) | 2021-06-22 | 2021-06-22 | Regional risk level assessment method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113298438A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116736780A (en) * | 2023-08-15 | 2023-09-12 | 贵州汇通华城股份有限公司 | Startup and shutdown control optimization method and system for regional energy station |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108154430A (en) * | 2017-12-28 | 2018-06-12 | 上海氪信信息技术有限公司 | A kind of credit scoring construction method based on machine learning and big data technology |
| CN110298577A (en) * | 2019-06-21 | 2019-10-01 | 济南大学 | Set disaster risk evaluating method and system for a kind of Yanhe Village based on DPSIR model |
| CN110377674A (en) * | 2019-06-13 | 2019-10-25 | 中国地质大学深圳研究院 | A kind of Typhoon Storm Surge Over methods of risk assessment and system based on ArcGIS platform |
| CN111738549A (en) * | 2020-05-21 | 2020-10-02 | 平安国际智慧城市科技股份有限公司 | Food safety risk assessment method, device, equipment and storage medium |
| CN111737651A (en) * | 2020-06-22 | 2020-10-02 | 黄河勘测规划设计研究院有限公司 | Spatial gridding drought disaster risk assessment method and system based on multi-source data |
| CN111950585A (en) * | 2020-06-29 | 2020-11-17 | 广东技术师范大学 | XGboost-based underground comprehensive pipe gallery safety condition assessment method |
| CN112287018A (en) * | 2020-11-06 | 2021-01-29 | 武汉理工大学 | Method and system for evaluating damage risk of 10kV tower under typhoon disaster |
-
2021
- 2021-06-22 CN CN202110694076.9A patent/CN113298438A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108154430A (en) * | 2017-12-28 | 2018-06-12 | 上海氪信信息技术有限公司 | A kind of credit scoring construction method based on machine learning and big data technology |
| CN110377674A (en) * | 2019-06-13 | 2019-10-25 | 中国地质大学深圳研究院 | A kind of Typhoon Storm Surge Over methods of risk assessment and system based on ArcGIS platform |
| CN110298577A (en) * | 2019-06-21 | 2019-10-01 | 济南大学 | Set disaster risk evaluating method and system for a kind of Yanhe Village based on DPSIR model |
| CN111738549A (en) * | 2020-05-21 | 2020-10-02 | 平安国际智慧城市科技股份有限公司 | Food safety risk assessment method, device, equipment and storage medium |
| CN111737651A (en) * | 2020-06-22 | 2020-10-02 | 黄河勘测规划设计研究院有限公司 | Spatial gridding drought disaster risk assessment method and system based on multi-source data |
| CN111950585A (en) * | 2020-06-29 | 2020-11-17 | 广东技术师范大学 | XGboost-based underground comprehensive pipe gallery safety condition assessment method |
| CN112287018A (en) * | 2020-11-06 | 2021-01-29 | 武汉理工大学 | Method and system for evaluating damage risk of 10kV tower under typhoon disaster |
Non-Patent Citations (1)
| Title |
|---|
| 弗朗索瓦•肖莱 等: "《R语言深度学习》", vol. 1, 31 March 2021, 机械工业出版社, pages: 209 - 211 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116736780A (en) * | 2023-08-15 | 2023-09-12 | 贵州汇通华城股份有限公司 | Startup and shutdown control optimization method and system for regional energy station |
| CN116736780B (en) * | 2023-08-15 | 2023-11-24 | 贵州汇通华城股份有限公司 | Startup and shutdown control optimization method and system for regional energy station |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| AU2018101946A4 (en) | Geographical multivariate flow data spatio-temporal autocorrelation analysis method based on cellular automaton | |
| Ferrier et al. | A new predictor of the irreplaceability of areas for achieving a conservation goal, its application to real-world planning, and a research agenda for further refinement | |
| Mosadeghi et al. | Comparison of Fuzzy-AHP and AHP in a spatial multi-criteria decision making model for urban land-use planning | |
| Pliscoff et al. | Effects of alternative sets of climatic predictors on species distribution models and associated estimates of extinction risk: A test with plants in an arid environment | |
| Feng et al. | Modeling dynamic urban growth using cellular automata and particle swarm optimization rules | |
| Yao et al. | Simulating urban growth processes by integrating cellular automata model and artificial optimization in Binhai New Area of Tianjin, China | |
| Chen et al. | Resilience assessment of regional areas against earthquakes using multi-source information fusion | |
| CN113159482A (en) | Method and system for evaluating information security risk | |
| US10984518B2 (en) | Methods and systems for assessing the quality of geospatial data | |
| CN111127062B (en) | Group fraud identification method and device based on space search algorithm | |
| Ranjbar et al. | A GIS-based approach for earthquake loss estimation based on the immediate extraction of damaged buildings | |
| KR102476663B1 (en) | System and method for comprehensive diagnosis of possible disasters in decayed urban regeneration areas | |
| Triantakonstantis et al. | A spatially heterogeneous expert based (SHEB) urban growth model using model regionalization | |
| González et al. | Dempster-Shafer theory of evidence: A new approach to spatially model wildfire risk potential in central Chile | |
| CN113298438A (en) | Regional risk level assessment method and device, computer equipment and storage medium | |
| CN112836590B (en) | Flood disaster monitoring method and device, electronic equipment and storage medium | |
| CN108038151B (en) | Monitoring area determining method and system | |
| Abid et al. | Present and Future of Artificial Intelligence in Disaster Management | |
| CN111178756B (en) | Multiple linear regression fire risk assessment method based on environmental big data | |
| Lagarias | Exploring land use policy scenarios with the use of a cellular automata-based model: urban sprawl containment and sustainable development in Thessaloniki | |
| Georgati et al. | Spatially explicit population projections: The case of Copenhagen, Denmark | |
| Mirakhorlo et al. | Integration of SimWeight and Markov Chain to Predict Land Use of Lavasanat Basin | |
| CN114358462A (en) | City safety risk assessment method and system and electronic equipment | |
| Seo et al. | Fire Risk Prediction Analysis Using Machine Learning Techniques. | |
| Moradi et al. | A novel approach to support majority voting in spatial group MCDM using density induced OWA operator for seismic vulnerability assessment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |