Training Method, Device, Equipment and Storage Medium of Land Attribute Classification Model

Abstract

An embodiment of the present application provides a training method, device, equipment and storage medium for a land use attribute classification model. The method includes: obtaining the longitude and latitude data marked with timestamps of mobile terminals within a preset geographical area, and the geographical fence data of the preset geographical area; the geographical fence data includes the longitude and latitude range data circled by each fence in the preset geographical area, and the land use attributes of each fence; determining the corresponding relationship between the longitude and latitude data marked with timestamps and the geographical fence data according to the longitude and latitude data marked with timestamps and the longitude and latitude range data to obtain associated data; dividing the associated data into a training set and a validation set according to a preset ratio; training a land use attribute classification model based on the training set and the validation set; the embodiment of the present application can solve the problem of low classification efficiency of the existing land use attribute classification method.

Description

Technical Field

[0001] This application belongs to the field of land use attribute classification, and particularly relates to a training method, device, equipment and storage medium for a land use attribute classification model. Background Art

[0002] With the continuous development and popularization of smart phones and information and communication technologies, large-scale trajectory data storage has become relatively common and has become an important source for mining user behavior patterns. Land use attributes are an important manifestation of user behavior patterns. For example, work land and residential land can be used to assist in the construction of smart cities, such as optimizing commuting routes, industrial layouts, analyzing population flow, etc., so as to reduce traffic congestion, improve the convenience and satisfaction of citizens' lives, etc.; it can also assist in the consulting decisions of industries such as commercial geography real estate and offline consumption, such as optimizing brand layouts and commercial positioning according to the residential and work location information of the consumer population, optimizing offline advertising placement, etc. However, the classification efficiency of existing land use attribute classification methods is relatively low. Summary of the Invention

[0003] Embodiments of this application provide a training method, device, equipment and storage medium for a land use attribute classification model, which can solve the problem of relatively low classification efficiency of existing land use attribute classification methods.

[0004] In a first aspect, an embodiment of this application provides a training method for a land use attribute classification model, including:

[0005] Obtain the longitude and latitude data marked with time stamps of mobile terminals within a preset geographical area, and the geographical fence data of the preset geographical area; the geographical fence data includes the longitude and latitude range data enclosed by each fence in the preset geographical area, and the land use attributes of each fence;

[0006] Determine the corresponding relationship between the longitude and latitude data marked with time stamps and the geographical fence data according to the longitude and latitude data marked with time stamps and the longitude and latitude range data to obtain associated data;

[0007] Divide the associated data into a training set and a validation set according to a preset ratio;

[0008] Train a land use attribute classification model based on the training set and the validation set.

[0009] In one embodiment, determining the corresponding relationship between the longitude and latitude data marked with time stamps and the geographical fence data according to the longitude and latitude data marked with time stamps and the longitude and latitude range data to obtain associated data includes:

[0010] Use a preset coding algorithm to convert the longitude and latitude data into first coding data, and convert the longitude and latitude range data enclosed by each fence into second coding data;

[0011] Taking the fact that the first encoded data falls within the data range represented by the second encoded data as the mapping relationship, determine the corresponding relationship among the first encoded data, the second encoded data, the land use attributes of each fence, and the timestamp to obtain the associated data.

[0012] In one embodiment, the training set includes M groups of first training sets, each group of first training sets includes the associated data of a fence in a preset geographical area, and the first training set is a training set with land use attributes of working land and residential land; the validation set includes N groups of first validation sets, each group of first validation sets includes the associated data of a fence in the preset geographical area, and the first validation set is a validation set with land use attributes of working land and residential land;

[0013] Train a land use attribute classification model based on the training set and the validation set, including:

[0014] Calculate the similarity degree of the distribution of longitude and latitude data in the first training set and the first validation set according to the timestamp to obtain M first parameters;

[0015] Determine the land use attribute with the most occurrences in the first training set corresponding to the first K first parameters with smaller values among the M first parameters as the predicted land use attribute of the first validation set;

[0016] Calculate the first hit rate that the predicted land use attribute of the first validation set is the same as the land use attribute of the first validation set to obtain N first hit rates;

[0017] Adjust the value of K until the N first hit rates meet the preset conditions to obtain the land use attribute classification model.

[0018] In one embodiment, training a land use attribute classification model based on the training set and the validation set further includes:

[0019] Calculate the overall similarity degree of the distribution of longitude and latitude data with the same land use attribute in the first collection according to the timestamp to obtain a second parameter; the first collection is the collection of the first training set and the first validation set;

[0020] Calculate the similarity degree of the distribution of longitude and latitude data in the second collection and the first collection according to the timestamp to obtain a third parameter; the second collection includes a preset proportion of the first training set, a preset proportion of the first validation set, a training set with land use attributes of other land, and a validation set with land use attributes of other land; other land is land other than residential land and working land;

[0021] Compare the second parameter and the third parameter through a preset comparison formula to obtain a comparison result;

[0022] Determine that the predicted land use attribute of the second set corresponding to the comparison result meeting the preset comparison condition is other land use, and determine that the predicted land use attribute of the second set corresponding to the comparison result not meeting the preset comparison condition is the land use attribute of the first set;

[0023] Calculate the second hit rate of the land use attribute of the second set and the predicted land use attribute of the second set;

[0024] Adjust the parameters of the preset comparison formula until the second hit rate meets the preset condition to obtain the land use attribute classification model.

[0025] In one embodiment, calculate the similarity degree of the distribution of longitude and latitude data in the first training set and the first validation set according to the time stamp to obtain M first parameters, including:

[0026] Calculate the empirical distribution functions of the first encoded data and the second encoded data respectively to obtain the first empirical distribution function curve and the second empirical distribution curve;

[0027] Calculate the area enclosed by the first empirical distribution function curve and the second empirical distribution curve to obtain M first parameters.

[0028] In a second aspect, an embodiment of the present application provides a method for classifying land use attributes using a land use attribute classification model. The land use attribute classification model is obtained by training through the training method of the land use attribute classification model. The method includes:

[0029] Obtain the communication feature data of the mobile terminal in the geographical area to be classified; the communication feature data is longitude and latitude data marked with time stamps;

[0030] Input the communication feature data into the land use attribute classification model for land use attribute classification, and output the land use attribute classification result of the geographical area to be classified.

[0031] In a third aspect, an embodiment of the present application provides a training device for a land use attribute classification model, including:

[0032] An acquisition module, configured to acquire the longitude and latitude data marked with time stamps of the mobile terminal in the preset geographical area, and the geographical fence data of the preset geographical area; the geographical fence data includes the longitude and latitude range data circled by each fence in the preset geographical area, and the land use attribute of each fence;

[0033] A determination module, configured to determine the corresponding relationship between the longitude and latitude data marked with time stamps and the geographical fence data according to the longitude and latitude data marked with time stamps and the longitude and latitude range data to obtain associated data;

[0034] A division module, configured to divide the associated data into a training set and a validation set according to a preset ratio;

[0035] A training module for training a land attribute classification model based on a training set and a validation set.

[0036] In one embodiment, the determination module includes a conversion unit and a determination unit;

[0037] The conversion unit is configured to convert the longitude and latitude data into first encoded data by using a preset encoding algorithm, and convert the longitude and latitude range data enclosed by each fence into second encoded data;

[0038] The determination unit is configured to determine the correspondence relationship among the first encoded data, the second encoded data, the land attributes of each fence, and the timestamp by using the mapping relationship that the first encoded data falls into the data interval represented by the second encoded data, so as to obtain associated data.

[0039] In one embodiment, the training set includes M groups of first training sets, each group of first training sets includes the associated data of a fence in a preset geographical area, and the first training set is a training set with land attributes of working land and residential land; the validation set includes N groups of first validation sets, each group of first validation sets includes the associated data of a fence in the preset geographical area, and the first validation set is a validation set with land attributes of working land and residential land;

[0040] The training module includes a calculation unit, a prediction unit, and an adjustment unit;

[0041] The calculation unit is configured to calculate the similarity degree of the distribution of the longitude and latitude data in the first training set and the first validation set according to the timestamp, so as to obtain M first parameters;

[0042] The prediction unit is configured to determine the predicted land attribute of the first validation set as the land attribute that appears most frequently in the first training set corresponding to the first K first parameters with smaller values among the M first parameters;

[0043] The calculation unit is further configured to calculate the first hit rate that the predicted land attribute of the first validation set is the same as the land attribute of the first validation set, so as to obtain N first hit rates;

[0044] The adjustment unit is configured to adjust the value of K until the N first hit rates meet the preset conditions, so as to obtain the land attribute classification model.

[0045] In one embodiment, the training module further includes a comparison unit;

[0046] The calculation unit is further configured to calculate the overall similarity degree of the distribution of the longitude and latitude data with the same land attribute in the first set according to the timestamp, so as to obtain a second parameter; the first set is the combination of the first training set and the first validation set;

[0047] A calculation unit is further configured to calculate the similarity degree of the distribution of longitude and latitude data in the second set and the first set according to timestamps, and obtain a third parameter; the second set includes a preset proportion of the first training set, a preset proportion of the first validation set, a training set with the land use attribute of other land, and a validation set with the land use attribute of other land; other land refers to land other than residential land and working land.

[0048] A comparison unit is configured to compare the second parameter and the third parameter through a preset comparison formula to obtain a comparison result.

[0049] A prediction unit is further configured to determine that the predicted land use attribute of the second set corresponding to the comparison result meeting the preset comparison condition is other land, and determine that the predicted land use attribute of the second set corresponding to the comparison result not meeting the preset comparison condition is the land use attribute of the first set.

[0050] A calculation unit is further configured to calculate the second hit rate of the land use attribute of the second set and the predicted land use attribute of the second set.

[0051] An adjustment unit is further configured to adjust the parameters of the preset comparison formula until the second hit rate meets the preset condition, and obtain a land use attribute classification model.

[0052] In one embodiment, the calculation unit is specifically configured to:

[0053] Calculate the empirical distribution functions of the first encoded data and the second encoded data respectively to obtain a first empirical distribution function curve and a second empirical distribution curve.

[0054] Calculate the area enclosed by the first empirical distribution function curve and the second empirical distribution curve to obtain M first parameters.

[0055] In a fourth aspect, an embodiment of the present application provides a device for classifying land use attributes using a land use attribute classification model. The land use attribute classification model is obtained by training with a training device for the land use attribute classification model. The device includes:

[0056] An acquisition module is configured to acquire communication feature data of a mobile terminal within a geographical area to be classified; the communication feature data is longitude and latitude data marked with timestamps.

[0057] A classification module is configured to input the communication feature data into the land use attribute classification model for land use attribute classification, and output a land use attribute classification result of the geographical area to be classified.

[0058] In a fifth aspect, an embodiment of the present application provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, the above method is implemented.

[0059] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium, on which an implementation program for information transmission is stored. When the program is executed by a processor, the above method is implemented.

[0060] For the training method, device, equipment and storage medium of the land use attribute classification model in the embodiment of the present application, the longitude and latitude data marked with timestamps of mobile terminals within a preset geographical area and the geographical fence data of the preset geographical area are obtained; the geographical fence data includes the longitude and latitude range data circled by each fence in the preset geographical area and the land use attributes of each fence; the corresponding relationship between the longitude and latitude data marked with timestamps and the geographical fence data is determined according to the longitude and latitude data marked with timestamps and the longitude and latitude range data to obtain associated data; the associated data is divided into a training set and a validation set according to a preset ratio; a land use attribute classification model is trained based on the training set and the validation set. The longitude and latitude data marked with timestamps can be obtained from the mobile terminal signaling, and the geographical fence data can be obtained from the operator of Location Based Services (LBS). That is to say, when training the model, the validation set does not need to be manually marked, and the land use attribute classification model can be trained, and the training efficiency of the land use attribute classification model is relatively high. When classifying the land use attributes based on the land use attribute classification model, the classification efficiency is relatively high. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required to be used in the embodiments of the present application. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0062] Figure 1 is a schematic flowchart of a method for training a land use attribute classification model provided by an embodiment of the present application;

[0063] Figure 2 is a schematic flowchart of a method for classifying land use attributes by applying a land use attribute classification model provided by an embodiment of the present application;

[0064] Figure 3 is a schematic structural diagram of a device for training a land use attribute classification model provided by an embodiment of the present application;

[0065] Figure 4 is a schematic structural diagram of a device for classifying land use attributes by applying a land use attribute classification model provided by an embodiment of the present application;

[0066] Figure 5 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0067] The features and exemplary embodiments of various aspects of the present application will be described in detail below. To make the objectives, technical solutions, and advantages of the present application clearer and more understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only configured to explain the present application and are not configured to limit the present application. For those skilled in the art, the present application can be implemented without some of these specific details. The following description of the embodiments is only provided to provide a better understanding of the present application by showing examples of the present application.

[0068] It should be noted that, in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the said element.

[0069] Currently, there are mainly three methods for classifying the land use attributes of plots: 1. Combining urban road network data and performing image recognition through satellite maps. This method is difficult to distinguish buildings with similar floors or appearances but different uses, and has low accuracy for points of interest (POIs) whose boundaries are not on the road network; 2. Conducting on-site surveys and manually marking. This method is very labor-intensive and difficult to cover the national map; 3. Making a land use classification map based on the planning documents issued by relevant government departments. This method requires a large amount of manpower, cannot be batch-processed due to different standards for planning documents issued in different regions, and is difficult to cover the national map. It can be seen that the classification efficiency of the existing land use attribute classification methods is relatively low.

[0070] To solve the problems of the prior art, an embodiment of the present application provides a training method, device, equipment, and storage medium for a land attribute classification model. In the embodiment of the present application, the longitude and latitude data marked with timestamps of a mobile terminal within a preset geographical area, and the geographical fence data of the preset geographical area are obtained; the geographical fence data includes the longitude and latitude range data circled by each fence in the preset geographical area, and the land attribute of each fence; the corresponding relationship between the longitude and latitude data marked with timestamps and the geographical fence data is determined according to the longitude and latitude data marked with timestamps and the longitude and latitude range data to obtain associated data; the associated data is divided into a training set and a validation set according to a preset ratio; a land attribute classification model is trained based on the training set and the validation set. The longitude and latitude data marked with timestamps can be obtained from the mobile terminal signaling, and the geographical fence data can be obtained from an operator of Location Based Services (LBS). That is to say, when training the model, there is no need for manual annotation of the validation set, and the land attribute classification model can be trained. The training efficiency of the land attribute classification model is relatively high, and the classification efficiency is relatively high when classifying land attributes based on the land attribute classification model. First, the training method for the land attribute classification model provided by the embodiment of the present application is introduced below.

[0071] Figure 1 FIG. 4 shows a schematic flowchart of a training method for a land attribute classification model provided by an embodiment of the present application. As Figure 1 shown, the method may include the following steps:

[0072] S110, obtain the longitude and latitude data marked with timestamps of a mobile terminal within a preset geographical area, and the geographical fence data of the preset geographical area.

[0073] Among them, the geographical fence data includes the longitude and latitude range data circled by each fence in the preset geographical area, and the land attribute of each fence. The longitude and latitude data marked with timestamps can be obtained from the mobile terminal signaling, and the geographical fence data can be obtained from an operator of Location Based Services (LBS).

[0074] S120, determine the corresponding relationship between the longitude and latitude data marked with timestamps and the geographical fence data according to the longitude and latitude data marked with timestamps and the longitude and latitude range data to obtain associated data.

[0075] In one embodiment, S120 may include:

[0076] Convert the longitude and latitude data into first encoded data by using a preset encoding algorithm, and convert the longitude and latitude range data circled by each fence into second encoded data;

[0077] Taking the fact that the first encoded data falls within the data range represented by the second encoded data as the mapping relationship, determine the corresponding relationship between the first encoded data, the second encoded data, the land use attributes of each fence, and the timestamp, to obtain associated data.

[0078] Among them, the preset encoding algorithm can convert longitude and latitude data into the first encoded data, and the longitude and latitude range data enclosed by each fence into the second encoded data. The preset encoding algorithm can be selected as GeoHash. GeoHash is essentially a way of spatial indexing. Its basic principle is to understand the earth as a two-dimensional plane, and recursively decompose the plane into smaller sub-blocks. Each sub-block has the same encoding within a certain longitude and latitude range. When GeoHash converts the above two types of data, the conversion granularity of the sub-blocks can be determined in advance based on the preset geographical area, so that the classification object of the model obtained by subsequent training corresponds to a geographical area of appropriate size. For example, if the area corresponding to the longitude and latitude data is the global area as the preset geographical area, the conversion granularity needs to be finer, and Geohash8 can be used for data conversion, that is, divide eight times in base 32 to obtain the above sub-blocks.

[0079] S130, divide the associated data into a training set and a validation set according to a preset ratio.

[0080] Among them, the preset ratio can be selected as 90% of the associated data for the training set and 10% of the associated data for the validation set.

[0081] S140, train a land use attribute classification model based on the training set and the validation set.

[0082] Among them, by training the model based on the training set and the validation set, a land use attribute classification model can be obtained. The algorithm used during training can be the k-NN algorithm.

[0083] In the embodiment of the present application, longitude and latitude data marked with timestamps of mobile terminals within a preset geographical area, and geographical fence data of the preset geographical area are obtained; the geographical fence data includes the longitude and latitude range data enclosed by each fence in the preset geographical area, and the land use attributes of each fence; according to the longitude and latitude data marked with timestamps and the longitude and latitude range data, determine the corresponding relationship between the longitude and latitude data marked with timestamps and the geographical fence data, to obtain associated data; divide the associated data into a training set and a validation set according to a preset ratio; train a land use attribute classification model based on the training set and the validation set. The longitude and latitude data marked with timestamps can be obtained from mobile terminal signaling, and the geographical fence data can be obtained from an operator of Location Based Services (LBS). That is to say, when training the model, there is no need for manual annotation of the validation set, and a land use attribute classification model can be trained. The training efficiency of the land use attribute classification model is relatively high, and the classification efficiency is relatively high based on the land use attribute classification model for land use attribute classification.

[0084] In one embodiment, the training set includes M groups of first training sets, each group of first training sets includes the associated data of a fence in a preset geographical area, and the first training set is a training set with land use attributes of working land and residential land; the validation set includes N groups of first validation sets, each group of first validation sets includes the associated data of a fence in the preset geographical area, and the first validation set is a validation set with land use attributes of working land and residential land; S140 may include:

[0085] S1401. Calculate the similarity degree of the distribution of longitude and latitude data in the first training set and the first validation set according to timestamps, and obtain M first parameters.

[0086] In one embodiment, S1401 may include:

[0087] Calculate the empirical distribution functions of the first encoded data and the second encoded data respectively to obtain a first empirical distribution function curve and a second empirical distribution curve; calculate the area enclosed by the first empirical distribution function curve and the second empirical distribution curve to obtain M first parameters.

[0088] Wherein, when calculating the empirical distribution function, using the number of the first encoded data and the second encoded data corresponding to each timestamp as variables, the first empirical distribution function curve and the second empirical distribution curve can be obtained, and the area enclosed by the first empirical distribution function curve and the second empirical distribution curve can be calculated. The smaller the area enclosed by the first empirical distribution function curve and the second empirical distribution curve, the greater the similarity degree of the distribution of longitude and latitude data in the first training set and the first validation set according to timestamps.

[0089] S1402. Determine the land use attribute with the most occurrences in the first training set corresponding to the first K first parameters with smaller values among the M first parameters as the predicted land use attribute of the first validation set.

[0090] Among them, the more times a certain land use attribute in the first training set corresponding to the first K first parameters appears, the greater the possibility that the land use attribute of the first validation set is the certain land use attribute of the first training set with more occurrences. K is a positive integer. For example, it can be selected from 1 to 10. The larger the value of K is selected, the higher the prediction accuracy is, but the amount of data to be processed for model training will increase accordingly.

[0091] S1403. Calculate the first hit rate that the predicted land use attribute of the first validation set is the same as the land use attribute of the first validation set, and obtain N first hit rates.

[0092] S1404. Adjust the value of K until the N first hit rates meet the preset conditions to obtain a land use attribute classification model.

[0093] Among them, the preset condition can be that a certain type of data of the overall situation of N first hit rates conforms to a certain numerical range. This certain type of data can be the average value, median, mode, etc., and this certain numerical range can be selected as 70%-100%. The larger the value of K is selected, the higher the prediction accuracy, but the corresponding data volume to be processed in model training will increase. Adjusting the value of K can ensure a relatively small data volume to be processed in the model training process while ensuring a relatively high precision of model training.

[0094] In an embodiment of the present application, timestamped longitude and latitude data of mobile terminals within a preset geographical area and geographical fence data of the preset geographical area are obtained; the geographical fence data includes the longitude and latitude range data enclosed by each fence in the preset geographical area and the land use attributes of each fence; the corresponding relationship between the timestamped longitude and latitude data and the geographical fence data is determined according to the timestamped longitude and latitude data and the longitude and latitude range data to obtain associated data; the associated data is divided into a training set and a validation set according to a preset ratio; a land use attribute classification model is trained based on the training set and the validation set. The timestamped longitude and latitude data can be obtained from mobile terminal signaling, and the geographical fence data can be obtained from an operator of Location Based Services (LBS). That is to say, when training the model, the validation set does not need to be manually labeled, and the land use attribute classification model can be trained, and the training efficiency of the land use attribute classification model is relatively high. When classifying land use attributes based on the land use attribute classification model, the classification efficiency is relatively high.

[0095] In one embodiment, S140: Training a land use attribute classification model based on a training set and a validation set may further include:

[0096] S1405, calculating the overall similarity degree of the distribution of longitude and latitude data with the same land use attribute in the first set according to the timestamp to obtain a second parameter.

[0097] Among them, the first set is the set of the first training set and the first validation set. The calculation process of the second parameter can refer to the calculation process of the first parameter in S1401. First, calculate the area enclosed by the empirical distribution curve of the longitude and latitude data with the same land use attribute in the first set. An empirical distribution curve enclosed area can be calculated for each fence, and the mean and standard deviation of the empirical distribution curve enclosed area are obtained to obtain the second parameter. Taking the first set as the set of land for residential use as an example, the second parameter can be the mean μ m and the standard deviation σ m . Taking the first set as the set of land for work use as an example, the second parameter can be the mean μ n and the standard deviation σ n .

[0098] S1406. Calculate the similarity degree of the distribution of longitude and latitude data in the second set and the first set according to the time stamps, and obtain the third parameter.

[0099] Among them, the second set includes a preset proportion of the first training set, a preset proportion of the first validation set, the training set with the land use attribute of other land uses, and the validation set with the land use attribute of other land uses; other land uses are land uses other than residential land and working land; the calculation process of the third parameter can refer to the calculation process of the second parameter in S1405, calculate the mean value of the area enclosed by the empirical distribution curves of the second set and the first set, and use this mean value as the third parameter.

[0100] S1407. Compare the second parameter and the third parameter through a preset comparison formula to obtain a comparison result.

[0101] Among them, the preset comparison formula for the second parameter corresponding to the land use attribute of the first set being residential land can be set as μ m +a×σ m ; the preset comparison formula for the second parameter corresponding to the land use attribute of the first set being working land can be set as μ n +b×σ n , where a and b are the parameters of the preset comparison formula.

[0102] S1408. Determine that the predicted land use attribute of the second set corresponding to the comparison result meeting the preset comparison condition is other land uses, and determine that the predicted land use attribute of the second set corresponding to the comparison result not meeting the preset comparison condition is the land use attribute of the first set.

[0103] Among them, taking the land use attribute of the first set being residential land as an example, the preset comparison condition can be that the third parameter is greater than the result calculated by the preset comparison formula.

[0104] S1409. Calculate the second hit rate of the land use attribute of the second set and the predicted land use attribute of the second set.

[0105] S1410. Adjust the parameters of the preset comparison formula until the second hit rate meets the preset conditions to obtain a land use attribute classification model.

[0106] Among them, different (a, b) combinations can be tried, for example, from 0 to 10 accurate to one decimal place, that is, a total of 10,000 combinations from (0.0, 0.0) to (10.0, 10.0), and select the combination with the highest second hit rate as the parameters of the land use attribute classification model.

[0107] The embodiments of the present application obtain the timestamped longitude and latitude data of mobile terminals within a preset geographical area, as well as the geographical fence data of the preset geographical area; the geographical fence data includes the longitude and latitude range data enclosed by each fence in the preset geographical area, and the land use attributes of each fence; determine the corresponding relationship between the timestamped longitude and latitude data and the geographical fence data according to the timestamped longitude and latitude data and the longitude and latitude range data to obtain associated data; divide the associated data into a training set and a validation set according to a preset ratio; train a land use attribute classification model based on the training set and the validation set. The timestamped longitude and latitude data can be obtained from the mobile terminal signaling, and the geographical fence data can be obtained from the operator of Location Based Services (LBS). That is to say, when training the model, there is no need for manual annotation of the validation set, and a land use attribute classification model can be trained, and the training efficiency of the land use attribute classification model is relatively high. Based on the land use attribute classification model, the land use attribute classification is carried out, and the classification efficiency is relatively high. Moreover, in the model training of the embodiments of the present application, the K-NN algorithm is referred to for model training. First, the model training with a secondary classification degree of land use attributes is carried out, and then the corresponding data of the third type of land use attribute is introduced, and the corresponding data of the third type of land use attribute is classified and trained by comparing the area mean values enclosed by the introduced empirical distribution curves. In this way, a two-stage training process is adopted. On the premise of taking into account the relatively simple operation process of the K-NN algorithm, the model training efficiency is improved, and the model prediction result is interpretable, ensuring that the trained land use attribute model has a relatively high classification accuracy.

[0108] The above embodiments introduce a training method for a land use attribute classification model provided by the embodiments of the present application. Based on this training method for the land use attribute classification model, the embodiments of the present application also provide a method for classifying land use attributes by using the land use attribute classification model. The land use attribute classification model is trained by this training method for the land use attribute classification model. The method may include:

[0109] S210, obtain the communication feature data of the mobile terminal within the geographical area to be classified.

[0110] Among them, the communication feature data is the timestamped longitude and latitude data.

[0111] S220, input the communication feature data into the land use attribute classification model for land use attribute classification, and output the land use attribute classification result of the geographical area to be classified.

[0112] Among them, the land use attribute classification result may include residential land, work land, and other land.

[0113] In the embodiment of the present application, the land use attribute classification model trained by using the above embodiment is applied to classify the land use attributes. Since the classification accuracy of the above land use attribute classification model is relatively high, it ensures that the accuracy of the land use attribute classification result obtained in the embodiment of the present application is relatively high.

[0114] Figure 1-2 The method provided by the embodiment of the present application is described. Next, the apparatus provided by the embodiment of the present application will be described with reference to the accompanying Figure 3-5 description of the apparatus provided by the embodiment of the present application.

[0115] Figure 3 FIG. shows a schematic structural diagram of a training apparatus for a land use attribute classification model provided by an embodiment of the present application. Figure 3 Each module in the shown apparatus has the function of implementing Figure 1 each step in and can achieve its corresponding technical effect. As Figure 3 shown, the apparatus may include:

[0116] An acquisition module 310, configured to acquire timestamped longitude and latitude data of mobile terminals within a preset geographical area, and geographical fence data of the preset geographical area; the geographical fence data includes longitude and latitude range data enclosed by each fence in the preset geographical area, and the land use attribute of each fence;

[0117] A determination module 320, configured to determine the correspondence between the timestamped longitude and latitude data and the geographical fence data according to the timestamped longitude and latitude data and the longitude and latitude range data, so as to obtain associated data;

[0118] A division module 330, configured to divide the associated data into a training set and a validation set according to a preset ratio;

[0119] A training module 340, configured to train a land use attribute classification model based on the training set and the validation set.

[0120] The embodiments of the present application obtain the longitude and latitude data marked with timestamps of mobile terminals within a preset geographical area, and the geographical fence data of the preset geographical area; the geographical fence data includes the longitude and latitude range data enclosed by each fence in the preset geographical area, and the land use attributes of each fence; determine the corresponding relationship between the longitude and latitude data marked with timestamps and the geographical fence data according to the longitude and latitude data marked with timestamps and the longitude and latitude range data to obtain associated data; divide the associated data into a training set and a validation set according to a preset ratio; train a land use attribute classification model based on the training set and the validation set. The longitude and latitude data marked with timestamps can be obtained from mobile terminal signaling, and the geographical fence data can be obtained from an operator of Location Based Services (LBS). That is to say, when training the model, there is no need for manual annotation of the validation set, and a land use attribute classification model can be trained. The training efficiency of the land use attribute classification model is relatively high, and the classification efficiency is relatively high based on the land use attribute classification model. Moreover, in the embodiments of the present application, when training the model, the K-NN algorithm is referred to for model training. First, model training is performed for the secondary classification degree of land use attributes, and then the corresponding data of the third type of land use attribute is introduced. By introducing the mean value of the area enclosed by the empirical distribution curve to classify and train the corresponding data of the third type of land use attribute, such a two-stage training process is adopted. On the premise of taking into account the relatively simple operation process of the K-NN algorithm, the model training efficiency is improved, and the model prediction result is interpretable, ensuring that the trained land use attribute model has a relatively high classification accuracy.

[0121] In one embodiment, the determination module 320 includes a conversion unit and a determination unit;

[0122] The conversion unit is used to convert the longitude and latitude data into first encoded data by using a preset encoding algorithm, and convert the longitude and latitude range data enclosed by each fence into second encoded data;

[0123] The determination unit is used to determine the corresponding relationship between the first encoded data, the second encoded data, the land use attributes of each fence, and the timestamp with the data interval represented by the first encoded data falling into the second encoded data as the mapping relationship to obtain associated data.

[0124] In one embodiment, the training set includes M groups of first training sets, each group of first training sets includes the associated data of a fence in the preset geographical area, and the first training set is a training set for land use attributes of working land and residential land; the validation set includes N groups of first validation sets, each group of first validation sets includes the associated data of a fence in the preset geographical area, and the first validation set is a validation set for land use attributes of working land and residential land;

[0125] The training module 340 includes a calculation unit, a prediction unit, and an adjustment unit;

[0126] A calculation unit, configured to calculate the similarity degree of the distribution of longitude and latitude data in the first training set and the first validation set according to timestamps, and obtain M first parameters;

[0127] A prediction unit, configured to determine the land use attribute with the most occurrences in the first training set corresponding to the first K first parameters with smaller values among the M first parameters as the predicted land use attribute of the first validation set;

[0128] The calculation unit is further configured to calculate the first hit rate that the predicted land use attribute of the first validation set is the same as the land use attribute of the first validation set, and obtain N first hit rates;

[0129] An adjustment unit, configured to adjust the value of K until the N first hit rates meet a preset condition, and obtain a land use attribute classification model.

[0130] In one embodiment, the training module 340 further includes a comparison unit;

[0131] The calculation unit is further configured to calculate the overall similarity degree of the distribution of longitude and latitude data with the same land use attribute in the first combined set according to timestamps, and obtain a second parameter; the first combined set is the combined set of the first training set and the first validation set;

[0132] The calculation unit is further configured to calculate the similarity degree of the distribution of longitude and latitude data in the second combined set and the first combined set according to timestamps, and obtain a third parameter; the second combined set includes a preset proportion of the first training set, a preset proportion of the first validation set, a training set with a land use attribute of other land, and a validation set with a land use attribute of other land; other land is land other than residential land and working land;

[0133] The comparison unit is configured to compare the second parameter and the third parameter through a preset comparison formula to obtain a comparison result;

[0134] The prediction unit is further configured to determine that the predicted land use attribute of the second combined set corresponding to the comparison result meeting the preset comparison condition is other land, and determine that the predicted land use attribute of the second combined set corresponding to the comparison result not meeting the preset comparison condition is the land use attribute of the first combined set;

[0135] The calculation unit is further configured to calculate the second hit rate of the land use attribute of the second combined set and the predicted land use attribute of the second combined set;

[0136] The adjustment unit is further configured to adjust the parameters of the preset comparison formula until the second hit rate meets a preset condition, and obtain a land use attribute classification model.

[0137] In one embodiment, the calculation unit is specifically configured to:

[0138] Calculate the empirical distribution functions of the first encoded data and the second encoded data respectively to obtain a first empirical distribution function curve and a second empirical distribution curve;

[0139] Calculate the area enclosed by the first empirical distribution function curve and the second empirical distribution curve to obtain M first parameters.

[0140] In the embodiments of the present application, the latitude and longitude data marked with timestamps of mobile terminals within a preset geographical area and the geographical fence data of the preset geographical area are obtained; the geographical fence data includes the latitude and longitude range data enclosed by each fence in the preset geographical area and the land use attributes of each fence; determine the corresponding relationship between the latitude and longitude data marked with timestamps and the geographical fence data according to the latitude and longitude data marked with timestamps and the latitude and longitude range data to obtain associated data; divide the associated data into a training set and a validation set according to a preset ratio; train a land use attribute classification model based on the training set and the validation set. The latitude and longitude data marked with timestamps can be obtained from the mobile terminal signaling, and the geographical fence data can be obtained from the operator of Location Based Services (LBS). That is to say, when training the model, there is no need for manual annotation of the validation set, and a land use attribute classification model can be trained. The training efficiency of the land use attribute classification model is relatively high. Based on the land use attribute classification model for land use attribute classification, the classification efficiency is relatively high. Moreover, in the embodiments of the present application, when training the model, refer to the K-NN algorithm for model training. First, train a model with a secondary classification degree of land use attributes, and then introduce the corresponding data of the third type of land use attribute and classify and train the corresponding data of the third type of land use attribute by introducing the comparison of the mean value of the area enclosed by the empirical distribution curve. In this way, a two-stage training process is adopted. On the premise of taking into account the relatively simple operation process of the K-NN algorithm, the model training efficiency is improved, and the model prediction result is interpretable, ensuring that the trained land use attribute model has a relatively high classification accuracy.

[0141] Figure 4 The structural schematic diagram of a device for classifying land use attributes by using a land use attribute classification model provided by an embodiment of the present application is shown. Figure 4 Each module in the shown device has the function of implementing Figure 2 each step in Figure 4 and can achieve its corresponding technical effects. As

[0142] An acquisition module 410, configured to acquire communication feature data of a mobile terminal within a geographical area to be classified; the communication feature data is latitude and longitude data marked with timestamps;

[0143] A classification module 420, configured to input the communication feature data into a land use attribute classification model for land use attribute classification and output a land use attribute classification result of the geographical area to be classified.

[0144] In the embodiments of the present application, the land use attribute classification model trained by the above embodiments is applied to classify land use attributes. Since the classification accuracy of the above land use attribute classification model is relatively high, it ensures that the accuracy of the land use attribute classification results obtained in the embodiments of the present application is relatively high.

[0145] Figure 5 The structural schematic diagram of an electronic device provided by an embodiment of the present application is shown. As Figure 5 shown, the device may include a processor 501 and a memory 502 storing computer program instructions.

[0146] Specifically, the above-mentioned processor 501 may include a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0147] The memory 502 may include a mass storage for data or instructions. By way of example and not limitation, the memory 502 may include a Hard Disk Drive (HDD), a floppy disk drive, a flash memory, an optical disc, a magneto-optical disc, a magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. In one example, the memory 502 may include removable or non-removable (or fixed) media, or the memory 502 is a non-volatile solid-state memory. The memory 502 may be inside or outside the integrated gateway disaster recovery device.

[0148] In one example, the memory 502 may be a Read Only Memory (ROM). In one example, the ROM may be a mask-programmed ROM, a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable PROM (EEPROM), an Electrically Rewritable ROM (EAROM), or a flash memory, or a combination of two or more of these.

[0149] The processor 501 reads and executes the computer program instructions stored in the memory 502 to implement Figure 1-2 the method in the embodiments shown, and achieves Figure 1-2 the corresponding technical effects achieved by the method executed by the instance shown. For the sake of brevity, it will not be described in detail here.

[0150] In one example, the electronic device may further include a communication interface 503 and a bus 510. Among them, as Figure 5As shown, a processor 501, a memory 502, and a communication interface 503 are connected via a bus 510 to complete communication with each other.

[0151] The communication interface 503 is mainly used to implement communication between various modules, devices, units, and / or equipment in the embodiments of the present application.

[0152] The bus 510 includes hardware, software, or both, and couples the components of the online data flow metering device to each other. By way of example and not limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses or a combination of two or more of these. In a suitable case, the bus 510 may include one or more buses. Although the embodiments of the present application describe and illustrate specific buses, the present application contemplates any suitable bus or interconnect.

[0153] The electronic device can execute the training method of the land use attribute classification model in the embodiments of the present application, so as to achieve Figure 1-2 the corresponding technical effects of the described training method of the land use attribute classification model.

[0154] In addition, in combination with the training method of the land use attribute classification model in the above embodiments, the embodiments of the present application can be implemented by providing a computer storage medium. Computer program instructions are stored on the computer storage medium; when the computer program instructions are executed by a processor, any one of the training methods of the land use attribute classification model in the above embodiments is implemented.

[0155] It should be clear that the present application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and illustrated as examples. However, the method process of the present application is not limited to the specific steps described and illustrated, and those skilled in the art can make various changes, modifications, and additions, or change the order between steps after understanding the spirit of the present application.

[0156] The functional blocks shown in the above-described structural block diagrams can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, and so on. When implemented in software, the elements of the present application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via a data signal carried in a carrier wave. A "machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROMs, flash memories, erasable ROMs (EROMs), floppy disks, CD-ROMs, optical discs, hard disks, fiber optic media, radio frequency (RF) links, and so on. The code segment can be downloaded via a computer network such as the Internet, an intranet, and so on.

[0157] It should also be noted that the exemplary embodiments mentioned in the present application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above steps, that is, the steps can be executed in the order mentioned in the embodiments, can be different from the order in the embodiments, or several steps can be executed simultaneously.

[0158] Aspects of the present application have been described above with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each block in the flowcharts and / or block diagrams, and the combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the functions / actions specified in one or more blocks of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each block in the block diagram and / or flowchart, and the combinations of blocks in the block diagram and / or flowchart, can also be implemented by dedicated hardware that performs the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0159] As described above, this is only the specific implementation manner of the present application. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, modules, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein. It should be understood that the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present application.

Claims

1. A training method for a land use attribute classification model, characterized in that, it includes: Obtain the longitude and latitude data marked with timestamps of mobile terminals within a preset geographical area, and the geographical fence data of the preset geographical area; The geographical fence data includes the longitude and latitude range data enclosed by each fence in the preset geographical area, and the land use attributes of each fence; Determine the correspondence between the longitude and latitude data marked with timestamps and the geographical fence data according to the longitude and latitude data marked with timestamps and the longitude and latitude range data to obtain associated data; Divide the associated data into a training set and a validation set according to a preset ratio; Train a land use attribute classification model based on the training set and the validation set.

2. The training method for a land use attribute classification model according to claim 1, characterized in that, The step of determining the correspondence between the longitude and latitude data marked with timestamps and the geographical fence data according to the longitude and latitude data marked with timestamps and the longitude and latitude range data to obtain associated data includes: Use a preset coding algorithm to convert the longitude and latitude data into first coding data, and convert the longitude and latitude range data enclosed by each fence into second coding data; Taking the mapping relationship that the first coding data falls into the data interval represented by the second coding data, determine the correspondence between the first coding data, the second coding data, the land use attributes of each fence, and the timestamp to obtain the associated data.

3. The training method for a land use attribute classification model according to claim 2, characterized in that, The training set includes M groups of first training sets, each group of first training sets includes the associated data of a fence in the preset geographical area, and the first training sets are the training sets with land use attributes of work land and residential land; the validation set includes N groups of first validation sets, each group of first validation sets includes the associated data of a fence in the preset geographical area, and the first validation sets are the validation sets with land use attributes of work land and residential land; The step of training a land use attribute classification model based on the training set and the validation set includes: Calculate the similarity degree of the distribution of the longitude and latitude data in the first training set and the first validation set according to the timestamp to obtain M first parameters; Determine the land use attribute that appears most frequently in the first training set corresponding to the first K first parameters with smaller values among the M first parameters as the predicted land use attribute of the first validation set; Calculate the first hit rate that the predicted land use attribute of the first validation set is the same as the land use attribute of the first validation set to obtain N first hit rates; Adjust the value of K until the N first hit rates meet the preset conditions to obtain the land use attribute classification model.

4. The training method for a land use attribute classification model according to claim 3, characterized in that, The step of training a land use attribute classification model based on the training set and the validation set further includes: Calculate the overall similarity of the distribution of the longitude and latitude data with the same land use attribute in the first set according to the time stamps to obtain a second parameter; the first set is the combination of the first training set and the first validation set; Calculate the similarity of the distribution of the longitude and latitude data in the second set and the first set according to the time stamps to obtain a third parameter; the second set includes a preset proportion of the first training set, a preset proportion of the first validation set, the training set with the land use attribute of other land uses, and the validation set with the land use attribute of other land uses; the other land uses are land uses other than residential land and working land; Compare the second parameter and the third parameter through a preset comparison formula to obtain a comparison result; Determine that the predicted land use attribute of the second set corresponding to the comparison result that meets the preset comparison condition is the other land use, and determine that the predicted land use attribute of the second set corresponding to the comparison result that does not meet the preset comparison condition is the land use attribute of the first set; Calculate the second hit rate of the land use attribute of the second set and the predicted land use attribute of the second set; Adjust the parameters of the preset comparison formula until the second hit rate meets the preset conditions to obtain the land use attribute classification model.

5. The training method of the land use attribute classification model according to claim 3 or 4, characterized in that, The calculation of the similarity of the distribution of the longitude and latitude data in the first training set and the first validation set according to the time stamps to obtain M first parameters includes: Calculate the empirical distribution functions of the first encoded data and the second encoded data respectively to obtain a first empirical distribution function curve and a second empirical distribution curve; Calculate the area enclosed by the first empirical distribution function curve and the second empirical distribution curve to obtain the M first parameters.

6. A method for classifying land use attributes by using a land use attribute classification model, characterized in that, The land use attribute classification model is trained by the training method of the land use attribute classification model according to claim 1, and the method includes: Obtain the communication feature data of the mobile terminal in the geographical area to be classified; the communication feature data is longitude and latitude data marked with time stamps; Input the communication feature data into the land use attribute classification model for land use attribute classification, and output the land use attribute classification result of the geographical area to be classified.

7. A training device for a land use attribute classification model, characterized in that, comprising: An acquisition module for acquiring the longitude and latitude data marked with time stamps of the mobile terminal in the preset geographical area, and the geographical fence data of the preset geographical area; The geographical fence data includes the longitude and latitude range data circled by each fence in the preset geographical area, and the land use attribute of each fence; A determination module for determining the corresponding relationship between the longitude and latitude data marked with time stamps and the geographical fence data according to the longitude and latitude data marked with time stamps and the longitude and latitude range data to obtain associated data; A division module for dividing the associated data into a training set and a validation set according to a preset ratio; A training module for training a land use attribute classification model based on the training set and the validation set.

8. The training device for the land use attribute classification model according to claim 7, wherein, the determination module includes a conversion unit and a determination unit; the conversion unit is configured to convert the longitude and latitude data into first encoded data by using a preset encoding algorithm, and convert the longitude and latitude range data circled by each fence into second encoded data; the determination unit is configured to determine the corresponding relationship among the first encoded data, the second encoded data, the land use attributes of each fence, and the timestamp with the mapping relationship that the first encoded data falls into the data interval represented by the second encoded data, so as to obtain the associated data.

9. The training device for the land use attribute classification model according to claim 8, wherein, the training set includes M groups of first training sets, each group of first training sets includes the associated data of a fence in the preset geographical area, and the first training set is the training set with the land use attributes of working land and residential land; the validation set includes N groups of first validation sets, each group of first validation sets includes the associated data of a fence in the preset geographical area, and the first validation set is the validation set with the land use attributes of working land and residential land; the training module includes a calculation unit, a prediction unit, and an adjustment unit; the calculation unit is configured to calculate the similarity degree of the distribution of the longitude and latitude data in the first training set and the first validation set according to the timestamp, so as to obtain M first parameters; the prediction unit is configured to determine the predicted land use attribute of the first validation set as the land use attribute that appears most frequently in the first training set corresponding to the first K first parameters with smaller values among the M first parameters; the calculation unit is further configured to calculate the first hit rate that the predicted land use attribute of the first validation set is the same as the land use attribute of the first validation set, so as to obtain N first hit rates; the adjustment unit is configured to adjust the value of K until the N first hit rates meet the preset conditions, so as to obtain the land use attribute classification model.

10. The training device for the land use attribute classification model according to claim 9, wherein, the training module further includes a comparison unit; the calculation unit is further configured to calculate the overall similarity degree of the distribution of the longitude and latitude data with the same land use attribute in the first set according to the timestamp, so as to obtain a second parameter; the first set is the set of the first training set and the first validation set; the calculation unit is further configured to calculate the similarity degree of the distribution of the longitude and latitude data in the second set and the first set according to the timestamp, so as to obtain a third parameter; the second set includes a preset proportion of the first training set, a preset proportion of the first validation set, the training set with the land use attribute of other land, and the validation set with the land use attribute of other land; the other land is the land other than residential land and working land; the comparison unit is configured to compare the second parameter and the third parameter through a preset comparison formula to obtain a comparison result; The prediction unit is further configured to determine that the predicted land use attribute of the second set corresponding to the comparison result meeting the preset comparison condition is the other land use, and determine that the predicted land use attribute of the second set corresponding to the comparison result not meeting the preset comparison condition is the land use attribute of the first set; The calculation unit is further configured to calculate a second hit rate of the land use attribute of the second set and the predicted land use attribute of the second set; The adjustment unit is further configured to adjust the parameters of the preset comparison formula until the second hit rate meets the preset condition, so as to obtain the land use attribute classification model.

11. The training device for the land use attribute classification model according to claim 9 or 10, wherein, The calculation unit is specifically configured to: Calculate the empirical distribution functions of the first encoded data and the second encoded data respectively, and obtain a first empirical distribution function curve and a second empirical distribution curve; Calculate the area enclosed by the first empirical distribution function curve and the second empirical distribution curve to obtain the M first parameters.

12. A device for classifying land use attributes by using a land use attribute classification model, wherein, The land use attribute classification model is obtained by training with the training device for the land use attribute classification model according to claim 7, and the device includes: An acquisition module, configured to acquire communication feature data of a mobile terminal within a geographical area to be classified; the communication feature data is longitude and latitude data marked with a time stamp; A classification module, configured to input the communication feature data into the land use attribute classification model for land use attribute classification, and output a land use attribute classification result of the geographical area to be classified.

13. An electronic device, wherein, It includes: A memory, a processor, and a computer program stored on the memory and executable on the processor, and when the computer program is executed by the processor, the method according to any one of claims 1 to 6 is implemented.

14. A computer-readable storage medium, wherein, An information transfer implementation program is stored on the computer-readable storage medium, and when the program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

15. A computer program product, wherein, When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device is caused to execute the method according to any one of claims 1 to 6.

Images