Processing method, device and processing equipment for flood susceptibility prediction model

Abstract

The invention provides a processing method, device and processing equipment for a flood susceptibility prediction model, provides a novel model training architecture, and provides a flood susceptibility prediction model through a single-sample data enhancement method based on one-dimensional multi-feature structured data and coupling application of a consistency regularization semi-supervised learning method. According to the method, the prediction result instability caused by the data problem is reduced, the model training can be well completed, the flood susceptibility prediction precision of the model is improved, and the method has good practical application value.

Description

Technical Field

[0001] The present application relates to the field of flood susceptibility prediction, and specifically to a method, device and processing equipment for a flood susceptibility prediction model. Background Art

[0002] Floods are one of the most common and destructive natural disasters in the world. Therefore, accurately identifying and predicting areas susceptible to heavy rains and floods is of great significance for preventing heavy rains and floods and mitigating disaster losses.

[0003] Flood susceptibility refers to the possibility of flood disasters occurring in a specific area under local topographic, hydrological and meteorological conditions. Flood susceptibility forecasting is a means of predicting flood-prone areas and can provide data support for flood response work, so as to respond to floods in advance or reduce the damage caused by floods.

[0004] The inventors of the present application have discovered that when using the currently widely used supervised learning method to predict flood susceptibility, the scarcity of flood data sets and incomplete data records will lead to uncertainty in the performance of the trained model.

[0005] Semi-supervised learning is a machine learning method between supervised learning and unsupervised learning. It is mainly used in situations where labeled data is scarce or sample labeling is difficult. In semi-supervised learning, the model can learn using a small amount of labeled data and a large amount of unlabeled data. Compared with supervised learning, semi-supervised learning methods can make full use of the information of unlabeled data, thereby achieving better generalization performance and higher prediction accuracy.

[0006] In semi-supervised learning methods, the use of unlabeled data can be divided into two paradigms: pseudo label (ProxyLabel) and consistency regularization (Consistency Regularization).

[0007] For the pseudo-label method, first, the pseudo-label method relies on the prediction results of the existing model. If the performance of the initial model is poor, the obtained pseudo-label may also be inaccurate, thus affecting the training and performance of the model; secondly, the pseudo-label method performs poorly in dealing with noise and erroneous labels because it cannot distinguish between real labels and pseudo-labels.

[0008] As for the consistency regularization method, it is a common semi-supervised learning regularization technique that aims to improve the generalization ability of the model by constraining the consistency between different inputs or different model predictions. Consistency regularization usually generates new samples by slightly perturbing unlabeled samples and requires these new samples to be consistent with the original samples in the model output. This perturbation technology usually uses data enhancement methods to enhance the data of a single sample and constrain the consistency with the original sample to improve the performance of the model.

[0009] In the fields of images, natural language, frequency signals, etc., semi-supervised learning methods based on consistency regularization have achieved remarkable results. However, in the field of flood susceptibility prediction, flood inventory datasets usually exist in a one-dimensional multi-feature structured form. Currently, data enhancement techniques for such datasets mainly target multiple samples, and there is a lack of data enhancement methods for a single sample. Due to this limitation, semi-supervised learning methods based on consistency regularization are difficult to be effectively applied in the field of flood susceptibility prediction. Summary of the invention

[0010] The present application provides a processing method, device and processing equipment for a flood susceptibility prediction model, and provides a novel model training architecture. Through a single sample data enhancement method based on one-dimensional multi-feature structured data, and coupled with an application consistency regularization semi-supervised learning method, the instability of prediction results caused by data problems is reduced, and model training can be completed more effectively, providing the model's flood susceptibility prediction accuracy, and has good practical application value.

[0011] In a first aspect, the present application provides a method for processing a flood susceptibility prediction model, the method comprising:

[0012] Acquire a flood inventory dataset, wherein the flood inventory dataset includes a first flood inventory dataset with a label and a second flood inventory dataset without a label, the label identifies whether a flood occurs in a corresponding flood inventory data portion, and the flood inventory dataset uses a quantitative value of a preset flood influencing factor as a feature;

[0013] Based on the k-means clustering method and the SMOTE method, single sample data enhancement is performed on the second flood inventory data set in the flood inventory data to obtain an enhanced data set;

[0014] Based on the flood inventory dataset and the enhanced dataset, the multi-layer perceptron model is trained with the consistency regularized semi-supervised learning method to obtain the flood susceptibility prediction model, which is used to predict the corresponding flood susceptibility based on the characteristic values ​​of flood influencing factors input into the model.

[0015] In a second aspect, the present application provides a processing device for a flood susceptibility prediction model, the device comprising:

[0016] an acquisition unit, configured to acquire a flood inventory data set, wherein the flood inventory data set includes a first flood inventory data set with a label and a second flood inventory data set without a label, the label identifies whether a flood occurs in a corresponding flood inventory data portion, and the flood inventory data set uses a quantitative value of a preset flood influencing factor as a feature;

[0017] An enhancement unit is used for performing single sample data enhancement on the second flood inventory data set in the flood inventory data based on a k-means clustering method and a SMOTE method to obtain an enhanced data set;

[0018] The training unit is used to train a multi-layer perceptron model based on the flood inventory dataset and the enhanced dataset in combination with a consistency regularized semi-supervised learning method to obtain a flood susceptibility prediction model, wherein the flood susceptibility prediction model is used to predict the corresponding flood susceptibility based on the characteristic values ​​of the flood influencing factors input into the model.

[0019] In a third aspect, the present application provides a processing device, including a processor and a memory, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the method provided in the first aspect of the present application or any possible implementation method of the first aspect of the present application is executed.

[0020] In a fourth aspect, the present application provides a computer-readable storage medium, which stores multiple instructions, and the instructions are suitable for a processor to load to execute the method provided by the first aspect of the present application or any possible implementation of the first aspect of the present application.

[0021] From the above content, it can be concluded that the present application has the following beneficial effects:

[0022] For the goal of flood susceptibility prediction based on consistency regularized semi-supervised learning, this application provides a novel model training architecture. Through a single sample data enhancement method based on one-dimensional multi-feature structured data and coupled with the application of a consistency regularized semi-supervised learning method, the instability of the prediction results caused by data problems is reduced, and the model training can be completed more effectively, providing the model's flood susceptibility prediction accuracy, which has good practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 A schematic flow chart of a method for processing a flood susceptibility prediction model for this application;

[0025] Figure 2 A schematic diagram of the architecture of the consistency regularized semi-supervised learning method for this application;

[0026] Figure 3 A schematic diagram showing the change of AUC value of the model in this application with the k value;

[0027] Figure 4 A schematic diagram of the structure of a processing device for the flood susceptibility prediction model of this application;

[0028] Figure 5 A schematic diagram of the structure of the processing equipment of this application. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0030] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or modules is not necessarily limited to those steps or modules clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices. The naming or numbering of steps in this application does not mean that the steps in the method flow must be executed in the time / logical sequence indicated by the naming or numbering. The process steps that have been named or numbered can change the execution order according to the technical purpose to be achieved, as long as the same or similar technical effects can be achieved.

[0031] The division of modules in this application is a logical division. There may be other division methods when it is implemented in actual applications. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection between modules can be electrical or other similar forms, which are not limited in this application. In addition, the modules or submodules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed in multiple circuit modules, and some or all of the modules may be selected according to actual needs to achieve the purpose of the present application.

[0032] Before introducing the processing method of the flood susceptibility prediction model provided by this application, the background content involved in this application is first introduced.

[0033] The processing method, device and computer-readable storage medium of the flood susceptibility prediction model provided in the present application can be applied to processing equipment and provide a novel model training architecture. Through a single sample data enhancement method based on one-dimensional multi-feature structured data and coupled with an application consistency regularized semi-supervised learning method, the instability of prediction results caused by data problems is reduced, and model training can be completed more effectively, providing the model's flood susceptibility prediction accuracy, which has good practical application value.

[0034] The processing method of the flood susceptibility prediction model mentioned in this application can be executed by a processing device of the flood susceptibility prediction model, or a server, physical host or user equipment (UE) and other different types of processing devices integrated with the processing device of the flood susceptibility prediction model. Among them, the processing device of the flood susceptibility prediction model can be implemented in hardware or software, and the UE can be a terminal device such as a smart phone, tablet computer, laptop computer, desktop computer or personal digital assistant (PDA), and the processing device can be set in the form of a device cluster.

[0035] It can be understood that the focus of the present application is on model training and processing, and the model training and processing is carried out on the basis of existing data. Therefore, the processing equipment that executes the processing method of the flood susceptibility prediction model of the present application, or the processing equipment equipped with the corresponding application service of the processing method of the flood susceptibility prediction model of the present application, in actual circumstances, only needs to meet the requirements of data processing capabilities, and its specific equipment type and equipment deployment form can be flexibly configured according to actual needs.

[0036] If the specific data collection work is involved, the processing equipment needs to be configured with corresponding data collection functions (involving the configuration of software and hardware).

[0037] Similarly, if the trained model is to be used in practice, the processing equipment must be configured according to the specific application scenario.

[0038] As an example, in this case, the processing device may specifically include a first device part for training the model in the background and a second device part for applying the model in the field.

[0039] Next, the processing method of the flood susceptibility prediction model provided by this application is introduced.

[0040] First, see Figure 1 , Figure 1 A schematic flow chart of a method for processing a flood susceptibility prediction model of the present application is shown. The method for processing a flood susceptibility prediction model provided by the present application may specifically include the following steps S101 to S103:

[0041] Step S101, obtaining a flood inventory dataset, wherein the flood inventory dataset includes a first flood inventory dataset with a label and a second flood inventory dataset without a label, the label identifies whether a flood occurs in a corresponding flood inventory data portion, and the flood inventory dataset uses a quantitative value of a preset flood impact factor as a feature;

[0042] It is understandable that in order to train the flood susceptibility prediction model, it is necessary to configure training samples for training. In this regard, the flood inventory dataset configured here is the initial training sample obtained. The initial training sample needs to be further processed in the future to obtain a better-performing overall training sample.

[0043] Among them, the acquisition and processing of the flood list dataset here, in actual situations, is usually the acquisition and processing of ready-made data, such as receiving a manually entered flood list dataset, extracting a flood list dataset from a local storage location, or downloading / receiving a flood list dataset from a certain device.

[0044] Of course, in some cases, the real-time collection and processing of flood inventory data sets is not ruled out. In this process, data processing work in different links may also be involved, such as the docking of the data source of the original data.

[0045] It should be understood that the model training architecture adopted in this application belongs to a consistency regularized semi-supervised learning method / architecture. For this, a flood list data set will be configured in which a label is used to mark whether a flood has occurred (for example, 1 indicates that there is a flood, and 0 indicates that there is no flood). There is also a flood list data set that is not marked by a label to indicate whether a flood has occurred. This situation belongs to the scope of the prior art, so it will not be described in detail here.

[0046] Among them, the essential content of the flood inventory dataset is based on the quantitative values ​​of preset flood impact factors as features, so it can also be understood by the feature set corresponding to the flood impact factors. This corresponds to the working principle of the flood susceptibility prediction model to be trained. The flood susceptibility prediction model is used to predict the corresponding flood susceptibility based on the characteristic values ​​of the flood impact factors input by the model.

[0047] It can be understood that the prediction and processing of flood susceptibility usually involves multiple and different flood influencing factors (variables, which use different numerical values ​​to represent the specific characteristics of the corresponding features). The specific flood influencing factors themselves can be directly configured manually, or directly searched by a machine through a search, or optimized through a search within the range of manually configured factors (i.e., the initial factor set). Considering that how to configure the flood influencing factors is not the focus of this application, the specific description will not be given here.

[0048] Step S102, based on the k-means clustering method and the SMOTE method, single sample data enhancement is performed on the second flood list data set in the flood list data to obtain an enhanced data set;

[0049] It is understandable that in the field of flood susceptibility prediction, in actual situations, flood inventory datasets usually exist in a one-dimensional multi-feature structured form, and the current data enhancement technology for such datasets mainly targets multiple samples, and lacks data enhancement methods for single samples. Due to this limitation, consistency regularized semi-supervised learning methods are difficult to be effectively applied in the field of flood susceptibility.

[0050] In this regard, the present application further processes the data based on the flood inventory data obtained previously to obtain training samples suitable for the consistency regularized semi-supervised learning method.

[0051] Specifically, this application focuses on the second flood inventory dataset that does not carry labels (used to mark whether floods have occurred). Through the coupled application of the k-means clustering method and the SMOTE method, a single sample data enhancement method is implemented to perform single sample data enhancement on the second flood inventory dataset to overcome the uncertainty problem in the model performance of the subsequently trained model due to the scarcity of data in the flood dataset and the one-sided data records. The obtained new flood inventory dataset can be recorded as an enhanced dataset.

[0052] The k-means clustering method involved in this process can also be called the k-means clustering algorithm (KMA), which is an existing clustering algorithm. The main processing content can be understood as follows:

[0053] Divide the data into K groups in advance, then randomly select K objects as the initial cluster centers, and then calculate the distance between each object and each seed cluster center, and assign each object to the cluster center closest to it. The cluster centers and the objects assigned to them represent a cluster. Each time a sample is assigned, the cluster center of the cluster will be recalculated based on the existing objects in the cluster. This process will be repeated until a certain termination condition is met. The termination condition can be that no (or a minimum number) of objects are reallocated to different clusters, no (or a minimum number) of cluster centers change again, and the sum of squared errors is locally minimized.

[0054] The SMOTE method involved, namely the Synthetic Minority Oversampling Technique, is an improved solution based on the random oversampling algorithm. Since random oversampling adopts a strategy of simply copying samples to increase minority class samples, it is easy to cause the problem of model overfitting, that is, the information learned by the model is too specific and not general enough. The basic idea of ​​the SMOTE algorithm is to analyze minority class samples and artificially synthesize new samples based on minority class samples and add them to the data set. The main processing content can be understood as:

[0055] (1) For each sample x in the minority class, calculate its distance to all samples in the minority class sample set using the Euclidean distance as the standard to obtain its k nearest neighbors; (2) Set a sampling ratio according to the sample imbalance ratio to determine the sampling multiplier N. For each minority class sample x, randomly select a number of samples from its k nearest neighbors. Suppose the selected nearest neighbor is o; (3) For each randomly selected nearest neighbor o, construct a new sample with the original sample according to the formula o(new) = o + rand(0,1)*(xo).

[0056] Step S103, based on the flood inventory dataset and the enhanced dataset, combined with the consistency regularized semi-supervised learning method, a multilayer perceptron model is trained to obtain a flood susceptibility prediction model, wherein the flood susceptibility prediction model is used to predict the corresponding flood susceptibility based on the characteristic values ​​of the flood influencing factors input into the model.

[0057] It can be understood that after the enhanced data set, which is a new training sample, is obtained through the previous step S102, it can be combined with the original flood inventory data set, which is a training sample, to form a whole, new training sample.

[0058] In this case, the model training logic of the consistency regularized semi-supervised learning method can be followed to carry out specific model training processing on the initial model specifically selected as the multilayer perceptron model, so as to obtain a flood susceptibility prediction model that has been trained and can be put into practical use.

[0059] Among them, the Multilayer Perceptron (MLP) model is a basic and widely used artificial neural network model. Its structure consists of multiple levels, including an input layer, at least one hidden layer, and an output layer. MLP has shown excellent performance in classification, regression, pattern recognition and other fields with its powerful nonlinear mapping capabilities and flexible structural design.

[0060] For model training processing, the following main processing contents may be involved:

[0061] In each round of model training, a group of training samples are input into the model so that the model can carry out the corresponding flood susceptibility prediction and realize forward propagation. Then, the loss function is calculated based on the flood susceptibility prediction results output by the model, and the model parameters are optimized according to the loss function calculation results to realize reverse propagation. In this way, when the model training requirements such as training time, training times or prediction accuracy are met, the model can be completed. At this time, the model is a flood susceptibility prediction model that can be put into practical use.

[0062] from Figure 1 It can be seen from the illustrated embodiments that, for the flood susceptibility prediction target based on consistency regularized semi-supervised learning, the present application provides a novel model training architecture, which reduces the instability of prediction results caused by data problems through a single sample data enhancement method based on one-dimensional multi-feature structured data and couples the application of a consistency regularized semi-supervised learning method. The model training can be completed more effectively, and the model's flood susceptibility prediction accuracy is provided, which has good practical application value.

[0063] Next, continue with the above Figure 1Each step of the illustrated embodiment and possible embodiments thereof in practical application are explained in detail.

[0064] As an exemplary embodiment, step S102 performs data enhancement on the second flood inventory data set in the flood inventory data based on the k-means clustering method and the SMOTE method to obtain an enhanced data set, which may specifically include:

[0065] The second flood inventory data set in the flood inventory data is clustered by using the k-means clustering method, and multiple cluster categories are obtained;

[0066] For multi-sample clustering categories, the SMOTE method is used to generate new samples between multiple samples included in the multi-sample clustering categories. At the same time, the samples included in the single-sample clustering categories are not processed, so as to achieve data enhancement of the flood inventory data and obtain an enhanced data set.

[0067] Among them, it can be understood that in the clustering process, when the clustering categories are determined, the samples with similarity based on the cluster centers will also be determined. These samples are the samples included in the clustering categories. In layman's terms, the clustering process is to divide the original data set into multiple clustering categories. Each category represents a specific data pattern or feature combination. The samples in each group have similar characteristics under a certain similarity measure.

[0068] There are two cases for clustering results. The first case is a multi-sample clustering category (the clustering category contains multiple samples), and the second case is a single-sample clustering category (the clustering category contains only one sample). This application is aimed at the multi-sample clustering category and performs data enhancement operations.

[0069] Specifically, for multi-sample clustering categories, the SMOTE method is used to generate new samples between multiple samples contained in the clustering category. These new samples inherit the characteristic distribution of the original samples, increase the diversity and sample size of the data set, and help improve the model training effect; for single-sample clustering categories, the data enhancement result is defined as the sample itself. In this case, due to the lack of sufficient samples to apply the SMOTE method, the original sample is retained as the output of data enhancement, which can ensure the robustness of the data enhancement process.

[0070] Furthermore, the following pseudo-code examples can help you understand:

[0071]

[0072] For the above pseudo code, we have:

[0073] The first line defines the method input, including: data set D (containing labeled and unlabeled data), the number of clusters K (that is, how many clusters the data set is expected to be divided into), and the parameter settings of the SMOTE algorithm;

[0074] The second line defines the method output, which is the data set after data augmentation, called the augmented data set D_augmented;

[0075] Lines 3 to 5 are the first step of the method, cluster analysis. The K-means algorithm is used to cluster the dataset D and obtain K clusters. The purpose of this step is to group the samples according to their similarities. For each cluster i, all samples in the cluster are extracted to form a sample set, which is recorded as sample set Si. In this way, each cluster corresponds to a sample set, which contains all the samples in the cluster.

[0076] The second step of the method is data augmentation. First, initialize an empty dataset augmented dataset D_augmented to store all new samples generated after data augmentation. This part is in line 6.

[0077] Lines 7-12 first traverse the sample set Si in each cluster i, and then perform data augmentation operations on each sample set. If the number of samples in the sample set Si is greater than 1, it means that there are enough samples for SMOTE data augmentation. In this case, the SMOTE method is applied to the sample set Si to generate a new sample set Ni, and the new sample set Ni generated by the SMOTE method is added to the enhanced dataset D_augmented, thereby expanding the dataset (lines 8-10). If there is only one sample in the sample set Si, the single sample in the sample set Si is directly added to the enhanced dataset D_augmented. In the case of only one sample, the SMOTE method cannot be applied, so the sample is directly retained (lines 11-12).

[0078] Finally, the enhanced dataset D_augmented after data augmentation is returned (Line 13).

[0079] As can be seen from the previous content, for the model training architecture of the consistency regularized semi-supervised learning method, this application has made certain changes and introduced data enhancement operations in order to use it for the one-dimensional multi-feature structured data involved in the application scenario of flood susceptibility prediction.

[0080] In short, the data enhancement of the original consistency regularized semi-supervised learning method uses the RandAugment and back-translation methods, which are used for data enhancement of image data and text data respectively.

[0081] In the present application, a single-sample data enhancement method is used to perform data enhancement on the flood susceptibility prediction dataset. During the data enhancement process, for a given single unlabeled sample input x, an enhanced data input x' is generated by the single-sample data enhancement method.

[0082] In addition, to more vividly understand the configuration work of the model training architecture corresponding to the consistency regularized semi-supervised learning method of this application, you can also refer to Figure 2 A schematic diagram of the architecture of the consistency regularized semi-supervised learning method of the present application is shown. Figure 2 It can be seen that this application has further optimized settings.

[0083] Specifically, as an exemplary embodiment, the consistency regularized semi-supervised learning method involved in this application may have the following configuration contents:

[0084] Use binary cross entropy loss to replace the original loss term;

[0085] The second flood inventory dataset and the enhanced dataset are predicted by the model respectively, and the output probability distribution of the two aspects is obtained. The distance between the output probability distribution of the two aspects is calculated by KL divergence as the unsupervised loss term;

[0086] The loss after model prediction of the first flood inventory dataset is used as the supervision loss item;

[0087] The overall loss is the sum of the supervised loss term and λ times the unsupervised loss term.

[0088] Specifically, for labeled data, the consistency regularized semi-supervised learning method uses traditional supervised losses, such as cross entropy loss, to ensure the performance of the model on labeled data. However, this application takes into account that the data used is binary data, so binary cross entropy loss is used to replace the original loss term.

[0089] In addition, for unlabeled data generated by data augmentation, the consistency loss of the model output between the original unlabeled samples and the augmented samples is calculated to encourage the model to maintain stable predictions for input changes, that is, the output of the two is calculated in the same model, and the distance between the output probability distributions of the two is calculated by KL divergence (Kullback-Leibler divergence) as the unsupervised loss term

[0090] On the other hand, for labeled samples, there is a normal supervision loss term;

[0091] In this case, the overall loss term is the sum of the supervised loss term and the unsupervised consistency loss term times λ.

[0092] During the training process, the model is trained by iteratively optimizing the overall loss function. In each iteration, the supervision loss on a batch of labeled data and the consistency loss corresponding to a batch of unlabeled data generated by data augmentation are calculated, and then the model parameters are updated to minimize these losses.

[0093] It can be seen that the embodiment here is how to achieve better model training effect. This application provides an optimized and novel architecture setting of the consistency regularized semi-supervised learning method.

[0094] In addition, for the initial model involved in this application, there is also a flood susceptibility prediction model obtained after the initial model training, that is, a multilayer perceptron model. As an exemplary embodiment, it can be specifically configured with a fully connected layer artificial neural network with 1 input layer, 2 hidden layers and 1 output layer. The number of nodes in each layer is 13, 128, 128 and 2, the activation function of the middle layer of the network is the ReLU activation function, and the activation function between the hidden layer and the output layer is the Sigmoid activation function.

[0095] It can be understood that a specific implementation solution is provided here for the model architecture involved in this application and the specific parameters / structures involved therein.

[0096] At the same time, the present application may also involve model setting work in terms of hyperparameters and other aspects for the flood susceptibility prediction model. These tasks can obviously be adjusted according to actual needs. Of course, they also include some routine operations. It can be understood that this application does not provide too much elaboration on the routine operations that can be adopted.

[0097] Taking the parameter adjustment process involved in hyperparameters as an example, as an exemplary embodiment, the configuration method of the flood susceptibility prediction model of the present application may also include:

[0098] Perform hyperparameter tuning for the k value of clustering categories and / or the weight λ of the unsupervised loss term involved in the consistency regularized semi-supervised learning method.

[0099] As for the clustering category k value, it can be understood that it is itself a parameter involved in the clustering process of the K-means clustering method, and this application can use it as a hyperparameter to carry out hyperparameter adjustment processing to facilitate better model training / prediction effects.

[0100] Among them, the hyperparameter k value can help select the best data enhancement data in single sample enhancement processing.

[0101] Similarly, in the embodiment here, the present application also treats the weight λ of the unsupervised loss term mentioned above as another hyperparameter to carry out hyperparameter adjustment processing to achieve better model training / prediction effects.

[0102] Among them, the hyperparameter λ can help to better carry out the back propagation link of the model and refer to the parameter adjustment of the model with higher efficiency / precision.

[0103] As an example here, taking a city's flood susceptibility dataset as an example, the number of labeled data samples in the dataset is 532, and the number of unlabeled data samples is 5000. After adjusting the weight of the unsupervised loss term, the model diverges when λ>1. After debugging, the final value of λ is 0.5, and the value of k is 20 to 3000. The model for each k value is trained 10 times, and its model AUC value is recorded. Figure 3 A schematic diagram showing the change of AUC value of the model of the present application with the k value is shown. Figure 3 In the figure, the box plot on the vertical axis is the AUC value of the model trained 10 times at the corresponding K value. The figure also shows the average AUC value of the model at each K value. The line connecting the average values ​​is the change curve. Finally, the box plot with the subscript x is the AUC value of the supervised learning MLP model.

[0104] Among them, the AUC value (Area Under Curve), which is defined as the area under the ROC curve and enclosed by the coordinate axis, is a commonly used indicator in model evaluation indicators.

[0105] In specific operations, the selection of the k value prioritizes the average accuracy of the AUC value, and secondly focuses on the stability of the model. Based on this, as an exemplary embodiment, the present application can specifically select 240 as the optimal value of the k value.

[0106] After testing on the test set, the AUC value of the model was 0.918, and the MLP accuracy of supervised learning training was 0.886. Compared with supervised learning, the performance of the model trained by this method was greatly improved.

[0107] After the model training is completed, it can obviously be put into practical use to predict the flood susceptibility of specific / specified areas.

[0108] In this regard, as an exemplary embodiment, the processing method of the flood susceptibility prediction model of the present application may also include:

[0109] Obtain the characteristic value of the target flood impact factor in the target area;

[0110] Inputting the characteristic value of the target flood influencing factor into the flood susceptibility prediction model, so that the flood susceptibility prediction model is based on the flood susceptibility corresponding to the prediction;

[0111] Extract the flood susceptibility prediction results of the target area output by the flood susceptibility prediction model.

[0112] It can be understood that in actual situations, the target area is usually determined according to the currently initiated flood susceptibility prediction task, and the initiation method of the flood susceptibility prediction task is flexible. It can be a manually initiated task, a task forwarded by other devices, or a task initiated autonomously according to a pre-configured task initiation strategy. These are all possible.

[0113] After the target area is determined, its corresponding flood impact factor characteristic value can be obtained, that is, the target flood impact factor characteristic value, which can also be referred to as a feature set. This data corresponds to the features involved in the training samples mentioned above.

[0114] In this way, after the characteristic value of the target flood influencing factor is input into the configured flood susceptibility prediction model, the flood susceptibility prediction result output after the model carries out the flood susceptibility prediction processing can be extracted.

[0115] At this point, the flood susceptibility prediction results can continue to be stored, displayed, forwarded, further processed or output to complete the flood susceptibility prompts, etc. The subsequent data application links of the flood susceptibility prediction results can obviously be flexibly configured and can be adjusted according to pre-configured application requirements and real-time application requirements. This application does not make specific limitations.

[0116] The above is an introduction to the processing method of the flood susceptibility prediction model provided by the present application. In order to facilitate better implementation of the processing method of the flood susceptibility prediction model provided by the present application, the present application also provides a processing device for the flood susceptibility prediction model from the perspective of functional modules.

[0117] See also Figure 4 , Figure 4 4 is a schematic diagram of a structure of a processing device for a flood susceptibility prediction model of the present application. In the present application, the processing device 400 for a flood susceptibility prediction model may specifically include the following structure:

[0118] The acquisition unit 401 is used to acquire a flood inventory dataset, wherein the flood inventory dataset includes a first flood inventory dataset with a tag and a second flood inventory dataset without a tag, the tag identifies whether a corresponding flood inventory data portion has a flood occurrence, and the flood inventory dataset uses a quantitative value of a preset flood impact factor as a feature;

[0119] The enhancement unit 402 is used to perform single sample data enhancement on the second flood inventory data set in the flood inventory data based on the k-means clustering method and the SMOTE method to obtain an enhanced data set;

[0120] The training unit 403 is used to train a multi-layer perceptron model based on the flood inventory dataset and the enhanced dataset in combination with a consistency regularized semi-supervised learning method to obtain a flood susceptibility prediction model, wherein the flood susceptibility prediction model is used to predict the corresponding flood susceptibility based on the characteristic values ​​of the flood influencing factors input into the model.

[0121] In an exemplary embodiment, the enhancement unit 402 is specifically configured to:

[0122] The second flood inventory data set in the flood inventory data is clustered by using the k-means clustering method, and multiple cluster categories are obtained;

[0123] For multi-sample clustering categories, the SMOTE method is used to generate new samples between multiple samples included in the multi-sample clustering categories. At the same time, the samples included in the single-sample clustering categories are not processed, so as to achieve data enhancement of the flood inventory data and obtain an enhanced data set.

[0124] In another exemplary embodiment, for the consistency regularized semi-supervised learning method, there are the following configuration contents:

[0125] Use binary cross entropy loss to replace the original loss term;

[0126] The second flood inventory dataset and the enhanced dataset are predicted by the model respectively, and the output probability distribution of the two aspects is obtained. The distance between the output probability distribution of the two aspects is calculated by KL divergence as the unsupervised loss term;

[0127] The loss after model prediction of the first flood inventory dataset is used as the supervision loss item;

[0128] The overall loss is the sum of the supervised loss term and λ times the unsupervised loss term.

[0129] In another exemplary embodiment, the multilayer perceptron model is a fully connected artificial neural network configured with 1 input layer, 3 hidden layers and 1 output layer, the number of nodes in each layer is 13, 128, 128 and 2, the activation function of the middle layer of the network is the ReLU activation function, and the activation function between the hidden layer and the output layer is the Sigmoid activation function.

[0130] In another exemplary embodiment, during the model training process, the apparatus further includes a parameter adjustment unit 404, which is used to:

[0131] Perform hyperparameter tuning for the k value of clustering categories and / or the weight λ of the unsupervised loss term involved in the consistency regularized semi-supervised learning method.

[0132] In yet another exemplary embodiment, after the hyperparameter tuning process, the clustering category k value is 240 and λ is 0.5.

[0133] In yet another exemplary embodiment, the apparatus further includes an application unit 405, configured to:

[0134] Obtain the characteristic value of the target flood impact factor in the target area;

[0135] Inputting the characteristic value of the target flood influencing factor into the flood susceptibility prediction model, so that the flood susceptibility prediction model is based on the flood susceptibility corresponding to the prediction;

[0136] Extract the flood susceptibility prediction results of the target area output by the flood susceptibility prediction model.

[0137] This application also provides a processing device from the perspective of hardware structure, see Figure 5 , Figure 5 501, a memory 502, and an input / output device 503. The processor 501 is used to execute the computer program stored in the memory 502 to implement the following Figure 1 The steps of the method for processing the flood susceptibility prediction model in the corresponding embodiment; or, the processor 501 is used to execute the computer program stored in the memory 502 to implement the following Figure 4 Corresponding to the functions of each unit in the embodiment, the memory 502 is used to store the processor 501 executing the above Figure 1 The computer program required for the processing method of the flood susceptibility prediction model in the corresponding embodiment.

[0138] Exemplarily, the computer program may be divided into one or more modules / units, one or more modules / units are stored in the memory 502, and executed by the processor 501 to complete the present application. One or more modules / units may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in the computer device.

[0139] The processing device may include, but is not limited to, a processor 501, a memory 502, and an input / output device 503. Those skilled in the art will appreciate that the illustration is merely an example of a processing device and does not constitute a limitation on the processing device, and may include more or fewer components than shown in the illustration, or a combination of certain components, or different components. For example, the processing device may also include a network access device, a bus, etc., and the processor 501, the memory 502, the input / output device 503, etc. are connected via a bus.

[0140] The processor 501 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the processing device, and uses various interfaces and lines to connect various parts of the entire device.

[0141] The memory 502 can be used to store computer programs and / or modules. The processor 501 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 502 and calling the data stored in the memory 502. The memory 502 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the processing device, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0142] When the processor 501 is used to execute the computer program stored in the memory 502, the following functions can be implemented:

[0143] Acquire a flood inventory dataset, wherein the flood inventory dataset includes a first flood inventory dataset with a label and a second flood inventory dataset without a label, the label identifies whether a flood occurs in a corresponding flood inventory data portion, and the flood inventory dataset uses a quantitative value of a preset flood influencing factor as a feature;

[0144] Based on the k-means clustering method and the SMOTE method, single sample data enhancement is performed on the second flood inventory data set in the flood inventory data to obtain an enhanced data set;

[0145] Based on the flood inventory dataset and the enhanced dataset, the multi-layer perceptron model is trained with the consistency regularized semi-supervised learning method to obtain the flood susceptibility prediction model, which is used to predict the corresponding flood susceptibility based on the characteristic values ​​of flood influencing factors input into the model.

[0146] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the processing device, processing equipment and corresponding units of the flood susceptibility prediction model described above can refer to the following. Figure 1 The description of the processing method of the flood susceptibility prediction model in the corresponding embodiment will not be repeated here in detail.

[0147] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0148] To this end, the present application provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions can be loaded by a processor to execute the present application as follows: Figure 1 The steps of the processing method of the flood susceptibility prediction model in the corresponding embodiment, the specific operation can refer to the following Figure 1 The description of the processing method of the flood susceptibility prediction model in the corresponding embodiment will not be repeated here.

[0149] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0150] Due to the instructions stored in the computer-readable storage medium, the present application can be executed. Figure 1 The steps of the processing method of the flood susceptibility prediction model in the corresponding embodiment, therefore, the present application can be implemented as follows Figure 1 The beneficial effects that can be achieved by the processing method of the flood susceptibility prediction model in the corresponding embodiment are detailed in the previous description and will not be repeated here.

[0151] The above is a detailed introduction to the processing method, device, processing equipment and computer-readable storage medium of the flood susceptibility prediction model provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for processing a flood susceptibility prediction model, characterized in that: The method comprises: Acquire a flood inventory dataset, wherein the flood inventory dataset includes a first flood inventory dataset carrying a tag and a second flood inventory dataset not carrying the tag, the tag identifies whether a flood occurs in a corresponding flood inventory data portion, and the flood inventory dataset uses a quantitative value of a preset flood impact factor as a feature; Based on the k-means clustering method and the SMOTE method, single sample data enhancement is performed on the second flood inventory data set in the flood inventory data to obtain an enhanced data set; Based on the flood inventory dataset and the enhanced dataset, a multilayer perceptron model is trained in combination with a consistency regularized semi-supervised learning method to obtain a flood susceptibility prediction model, wherein the flood susceptibility prediction model is used to predict the corresponding flood susceptibility based on the characteristic values ​​of the flood influencing factors input into the model.

2. The method according to claim 1, characterized in that: The method of performing data enhancement on the second flood inventory data set in the flood inventory data based on the k-means clustering method and the SMOTE method to obtain an enhanced data set includes: Clustering the second flood inventory data set in the flood inventory data by using the k-means clustering method to obtain a plurality of cluster categories; For the multi-sample clustering category, the SMOTE method is used to generate new samples between the multiple samples included in the multi-sample clustering category, and the samples included in the single-sample clustering category are not processed, so as to achieve data enhancement of the flood inventory data and obtain the enhanced data set.

3. The method according to claim 1, characterized in that For the consistency regularized semi-supervised learning method, the following configuration contents are included: Use binary cross entropy loss to replace the original loss term; The second flood inventory dataset and the enhanced dataset are respectively subjected to model prediction to obtain output probability distributions in two aspects, and the distance between the output probability distributions in the two aspects is calculated by KL divergence as an unsupervised loss term; The loss after the model prediction of the first flood inventory dataset is used as the supervision loss item; The overall loss is the sum of the supervised loss term and λ times the unsupervised loss term.

4. The method according to claim 1, characterized in that: The multilayer perceptron model is a fully connected artificial neural network configured with 1 input layer, 2 hidden layers and 1 output layer. The number of nodes in each layer is 13, 128, 128 and 2, the activation function of the middle layer of the network is the ReLU activation function, and the activation function between the hidden layer and the output layer is the Sigmoid activation function.

5. The method according to claim 1, characterized in that During the model training process, the method further includes: A hyperparameter adjustment process is performed for the clustering category k value and / or the weight λ of the unsupervised loss term involved in the consistency regularized semi-supervised learning method.

6. The method according to claim 5, characterized in that After the hyperparameter adjustment process, the cluster category k value is 240 and λ is 0.

5.

7. The method according to claim 1, characterized in that The method further comprises: Obtain the characteristic value of the target flood impact factor in the target area; Inputting the target flood impact factor characteristic value into the flood susceptibility prediction model, so that the flood susceptibility prediction model is based on the corresponding flood susceptibility predicted; The flood susceptibility prediction result of the target area output by the flood susceptibility prediction model is extracted.

8. A processing device for a flood susceptibility prediction model, characterized in that: The device comprises: an acquisition unit, configured to acquire a flood inventory data set, wherein the flood inventory data set includes a first flood inventory data set carrying a tag and a second flood inventory data set not carrying the tag, the tag identifies whether a flood occurs in a corresponding flood inventory data portion, and the flood inventory data set uses a quantitative value of a preset flood impact factor as a feature; an enhancement unit, configured to perform single sample data enhancement on the second flood inventory data set in the flood inventory data based on a k-means clustering method and a SMOTE method to obtain an enhanced data set; A training unit is used to train a multi-layer perceptron model based on the flood inventory dataset and the enhanced dataset in combination with a consistency regularized semi-supervised learning method to obtain a flood susceptibility prediction model, wherein the flood susceptibility prediction model is used to predict the corresponding flood susceptibility based on the characteristic values ​​of the flood influencing factors input into the model.

9. A processing device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program, and the processor executes the method according to any one of claims 1 to 7 when calling the computer program in the memory.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the method according to any one of claims 1 to 7.

Images