Object vehicle dispatching prediction model determination method and device, equipment, medium and product
By classifying the object car delivery training data and adjusting the initial model parameters, the problems of poor fitting ability and inapplicability of individuals in the existing technology are solved, and a more efficient object car delivery prediction model is achieved.
Patent Information
- Application Number
- CN202411978222.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-30
- Publication Date
- 2025-04-29
AI Technical Summary
The existing technology has problems such as poor fitting ability and inapplicability of individuals when predicting the object to leave the vehicle. It is mainly because drivers are not considered and drivers with different service abilities are placed in a strategy and model for prediction, resulting in the model being unable to fit huge data with huge differences, resulting in a decrease in the overall estimate effect.
By obtaining the object's outbound training data set, classifying the first object, obtaining the target number of classification categories, inputting the training data under each classification category to the corresponding initial model, and adjusting the preset parameters of the initial model to obtain the target model.
The fitting ability of the target model is enhanced, the applicability of the target model to objects is improved, and the problems of poor fitting ability and inapplicability of individuals are solved.
Smart Images
Figure CN119903435A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of big data analysis and machine learning, and in particular to a method, device, equipment, medium and product for determining a vehicle dispatch prediction model. Background Art
[0002] With the booming development of the online ride-hailing industry, estimating the probability of drivers going out can not only improve operational efficiency, but also provide data support for improving user experience and enhancing market competitiveness.
[0003] To address the above issues, related technologies usually use strategy-based and time series analysis-based methods to estimate driver behavior. However, the above methods have the following problems: Since they do not take into account the differences among drivers, all drivers with different service capabilities are estimated in one strategy and model. The model cannot fit the data with huge differences, resulting in a decrease in the overall estimation effect. In addition, since they do not take into account the differences among drivers, the above methods usually make inaccurate estimates in the individual driver dimension, and the estimation results cannot be used in scenarios at the individual driver level.
[0004] Therefore, the related technology has the problems of poor fitting ability and individual inapplicability when predicting the departure of an object. Summary of the invention
[0005] In view of this, the present disclosure provides a method, device, equipment, medium and product for determining an object vehicle departure prediction model to solve the problems of poor fitting ability and individual inapplicability of related technologies when predicting object vehicle departure.
[0006] In a first aspect, the present disclosure provides a method for determining a vehicle departure prediction model of an object, the method comprising:
[0007] Obtain a first object vehicle dispatching training data set, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object vehicle dispatching behavior, and the sample label is used to characterize the first object vehicle dispatching specific behavior;
[0008] Classify the first object according to the training data to obtain a target number of classification categories;
[0009] Input the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result;
[0010] According to the initial prediction results and sample labels, the preset parameters of the initial model are adjusted to obtain the target model.
[0011] In the disclosed embodiment, a training data set for the first object's vehicle dispatch is obtained, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object's vehicle dispatch behavior, and the sample label is used to characterize the first object's specific vehicle dispatch behavior; the first object is classified according to the training data to obtain a target number of classification categories; the training data of the first object under each classification category is input into the corresponding initial model to obtain an initial prediction result; according to the initial prediction result and the sample label, the preset parameters of the initial model are adjusted to obtain a target model. Since the disclosed embodiment classifies the first object according to the difference in the training data, and inputs the training data of the first object under each classification category into the corresponding initial model, and adjusts the preset parameters of the initial model to obtain the target model, the fitting ability of the target model is enhanced, and the applicability of the target model to the first object is improved.
[0012] In an optional implementation, before classifying the first object according to the training data to obtain a target number of classification categories, the method further includes:
[0013] Preprocess the training data to obtain the eigenvalues;
[0014] The first object is classified according to the feature value to obtain a target number of classification categories.
[0015] In the embodiment of the present disclosure, by performing data preprocessing on the training data and converting the training data into feature values, the quality and consistency of the training data can be ensured.
[0016] In an optional implementation, the first object is classified according to the training data to obtain a target number of classification categories, including:
[0017] According to the business service data of the first object, obtaining a first sub-dataset and a second sub-dataset from the training data set;
[0018] Obtaining a first number of classification categories according to the first training data in the first sub-dataset and a preset algorithm;
[0019] According to the first object dispatching service platform in the second sub-dataset, a second number of classification categories are obtained;
[0020] According to the first number of classification categories and the second number of classification categories, a target number of classification categories is obtained.
[0021] In the disclosed embodiment, by classifying the first object according to the first training data and the preset algorithm, a first number of classification categories are obtained, and the first object is classified in a first dimension. By classifying the first object according to the first object's vehicle dispatch service platform, a second number of classification categories are obtained, and the first object is classified in a second dimension. By obtaining a target number of classification categories according to the first number of classification categories and the second number of classification categories, the first object is classified in two different dimensions according to the differences of the first object in the training data.
[0022] In an optional implementation, before inputting the training data of the first object under each classification category into the corresponding initial model to obtain the initial prediction result, the method further includes:
[0023] Obtaining preset parameters, wherein the preset parameters are used to adjust the influence of the characteristic value;
[0024] Construct the corresponding initial model based on the feature values, preset parameters and current classification categories.
[0025] In the embodiment of the present disclosure, by constructing a corresponding initial model according to the characteristic values, preset parameters and the current classification category, the input training data can be processed to obtain an initial prediction result.
[0026] In an optional implementation, inputting the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result includes:
[0027] Input the training data of the first object under each classification category into the corresponding initial model to obtain an output result;
[0028] An initial prediction result is obtained according to the output result and a preset function, wherein the preset function is used to convert the output result into a probability value, and the probability value is used to characterize the prediction result corresponding to the training data of the first object.
[0029] In the disclosed embodiment, by inputting the training data of the first object under each classification category into the corresponding initial model for training, the fitting ability of the initial model can be enhanced and the applicability of the initial model for the first object can be improved. By obtaining the initial prediction result according to the output result and the preset function, the output result of the initial model can be converted into the initial prediction result.
[0030] In an optional implementation, after adjusting preset parameters of a corresponding target number of initial models according to the initial prediction results and the sample labels to obtain a target number of target models, the method further includes:
[0031] Acquire target data of a second object for which the vehicle departure behavior is to be predicted;
[0032] Based on the target number of classification categories, determine the target classification category to which the target data belongs;
[0033] Input the target data into the target model corresponding to the target classification category to obtain the target prediction result;
[0034] obtaining a target behavior of a second object;
[0035] According to the target prediction result and the target behavior, the target parameters of the corresponding target model are adjusted to obtain the adjusted target model.
[0036] In the disclosed embodiment, by determining the target classification category to which the target data belongs based on the target number classification categories, the target data can be input into the target model corresponding to the target classification category according to the difference of the target data to obtain the target prediction result, thereby enhancing the applicability of the target model to the second object. By adjusting the target parameters of the corresponding target model according to the target prediction result and the target behavior, the target model can be optimized to obtain the adjusted target model.
[0037] In a second aspect, the present disclosure provides a device for determining a vehicle departure prediction model of an object, the device comprising:
[0038] A first acquisition module is used to acquire a first object vehicle dispatching training data set, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object vehicle dispatching behavior, and the sample label is used to characterize the first object vehicle dispatching specific behavior;
[0039] A first obtaining module is used to classify the first object according to the training data to obtain a target number of classification categories;
[0040] A second obtaining module is used to input the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result;
[0041] The third module is used to adjust the preset parameters of the initial model according to the initial prediction results and sample labels to obtain the target model.
[0042] In a third aspect, the present disclosure provides a computer device, including: a memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for determining the object vehicle departure prediction model of the above-mentioned first aspect or any corresponding embodiment thereof by executing the computer instructions.
[0043] In a fourth aspect, the present disclosure provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method for determining an object vehicle departure prediction model of the first aspect or any corresponding embodiment thereof.
[0044] In a fifth aspect, the present disclosure provides a computer program product, including computer instructions, which are used to enable a computer to execute the method for determining an object vehicle departure prediction model according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] In order to more clearly illustrate the specific embodiments of the present disclosure or the technical solutions in the prior art, the drawings required for use in the specific embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0046] Figure 1 is a flowchart of a method for determining a vehicle departure prediction model according to an embodiment of the present disclosure;
[0047] Figure 2 is a structural block diagram of a device for determining an object vehicle departure prediction model according to an embodiment of the present disclosure;
[0048] Figure 3 It is a schematic diagram of the hardware structure of the computer device of the embodiment of the present disclosure. DETAILED DESCRIPTION
[0049] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present disclosure.
[0050] It is understandable that before using the technical solutions disclosed in the various embodiments of the present disclosure, the types, scopes of use, usage scenarios, etc. of the information involved in the present disclosure should be informed to relevant users and their authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations. The relevant users may include any type of right holders, such as individuals, enterprises, and groups.
[0051] For example, in response to receiving an active request from a user, a prompt message is sent to the relevant user to clearly prompt the relevant user that the operation requested to be performed will require obtaining and using the information of the relevant user, so that the relevant user can independently choose whether to provide information to software or hardware such as an electronic device, application, server or storage medium that executes the operation of the technical solution of the present disclosure based on the prompt message.
[0052] As an optional but non-limiting implementation, in response to receiving an active request from a relevant user, a prompt message is sent to the relevant user, for example, in the form of a pop-up window, in which the prompt message may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to select "agree" or "disagree" to provide information to the electronic device.
[0053] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet relevant laws and regulations may also be applied to the implementation of the present disclosure.
[0054] With the booming development of the online ride-hailing industry, estimating the probability of drivers going out can not only improve operational efficiency, but also provide data support for improving user experience and enhancing market competitiveness.
[0055] To address the above issues, related technologies usually use strategy-based and time series analysis-based methods to estimate driver behavior. However, the above methods have the following problems: Since they do not take into account the differences among drivers, all drivers with different service capabilities are estimated in one strategy and model. The model cannot fit the data with huge differences, resulting in a decrease in the overall estimation effect. In addition, since they do not take into account the differences among drivers, the above methods usually make inaccurate estimates in the individual driver dimension, and the estimation results cannot be used in scenarios at the individual driver level.
[0056] Therefore, the relevant technologies have the problems of poor fitting ability and individual inapplicability when predicting the vehicle departure of the object.
[0057] In order to solve the above problems, according to an embodiment of the present disclosure, an embodiment of a method for determining an object vehicle departure prediction model is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0058] In this embodiment, a method for determining an object vehicle departure prediction model is provided, such as Figure 1 As shown, Figure 1 : is a flow chart of a method for determining a vehicle departure prediction model according to an embodiment of the present disclosure. The flow chart can be applied to a server and includes the following steps:
[0059] Step S101, obtaining a first object vehicle dispatching training data set, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object vehicle dispatching behavior, and the sample label is used to characterize the first object vehicle dispatching specific behavior.
[0060] Optionally, in the disclosed embodiment, the first object is a historical driver of the online car-hailing platform, and the training data is used to predict the historical driver's driving behavior, including basic attribute information, driver's historical behavior, time-related characteristics, geographic location information, environmental characteristics, and platform incentive activities. The sample label is used to characterize the specific behavior of the historical driver's driving (including driving and not driving), and the sample label includes 0 and 1.
[0061] Specifically, the server can obtain training data through the service logs of the online car-hailing platform, user feedback, interaction records between drivers and passengers, third-party meteorological data and traffic data, etc.
[0062] It should be noted that the basic attribute information includes basic driver information and vehicle information, among which the basic driver information includes driver identification, driving experience, gender, age and driver's license type, and the vehicle information includes vehicle model, age, vehicle type (such as fuel vehicle, electric vehicle) and vehicle maintenance record, etc.; the driver's historical behavior includes historical dispatch frequency, historical order completion volume, historical dispatch time period (such as peak period, non-peak period), historical dispatch time, average per-order revenue, total revenue and revenue fluctuation; time-related characteristics include days of the week (such as weekdays, weekends), time periods of the day (such as morning, afternoon, evening), holidays and special events (such as concerts, sporting events); geographic location information includes regular dispatch areas, popularity of dispatch areas and supply and demand conditions in dispatch areas; environmental characteristics include weather conditions and traffic conditions, among which weather conditions include temperature, rainfall probability, wind speed and visibility, and traffic conditions include traffic congestion index, etc.; platform incentive activities include various commission-free activities, rush order activities and time rewards, etc.
[0063] Step S102: classify the first object according to the training data to obtain a target number of classification categories.
[0064] Optionally, in the disclosed embodiment, the server may classify historical drivers in different dimensions according to the differences in training data to obtain a target number of classification categories.
[0065] Step S103: input the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result.
[0066] Optionally, in the disclosed embodiment, the initial model may adopt a logistic regression model (LR). The initial prediction result is a prediction result of the historical driver's driving behavior.
[0067] Specifically, logistic regression is a statistical method for classification problems, mainly used for binary classification (i.e., the result is 0 or 1). Its core idea is to use the features of linear combination to predict the probability of an event, and map the probability to between 0 and 1 through a logical function.
[0068] Step S104, adjusting the preset parameters of the initial model according to the initial prediction results and sample labels to obtain the target model.
[0069] Optionally, in the embodiment of the present disclosure, the preset parameters refer to the parameters of the logistic regression model, including the number of consecutive days without the vehicle being dispatched, the number of consecutive days of dispatching the vehicle, the number of days with completed orders in the past 7 days, the average daily number of completed orders in the past 7 days, the average daily dispatch time in the past 7 days, guaranteed activity incentives, and order-rushing activity incentives and other parameters.
[0070] Specifically, the server can use an optimization algorithm (such as gradient descent, Newton method or quasi-Newton method) to adjust the preset parameters of the logistic regression model according to the initial prediction results and sample labels to minimize the loss function, obtain the optimal parameters, and then obtain the target model. For example, among the optimal parameters obtained in this embodiment, the parameter of the rush order activity is 0.1206, which is significantly positive, indicating that the rush order activity may promote the historical driver to drive; the parameter of the guaranteed minimum activity is about 0, indicating that the impact of the guaranteed minimum activity on whether the historical driver drives is not significant.
[0071] In the disclosed embodiment, a training data set for the first object's vehicle dispatch is obtained, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object's vehicle dispatch behavior, and the sample label is used to characterize the first object's specific vehicle dispatch behavior; the first object is classified according to the training data to obtain a target number of classification categories; the training data of the first object under each classification category is input into the corresponding initial model to obtain an initial prediction result; according to the initial prediction result and the sample label, the preset parameters of the initial model are adjusted to obtain a target model. Since the disclosed embodiment classifies the first object according to the difference in the training data, and inputs the training data of the first object under each classification category into the corresponding initial model, and adjusts the preset parameters of the initial model to obtain the target model, the fitting ability of the target model is enhanced, and the applicability of the target model to the first object is improved.
[0072] In some optional implementations, before classifying the first object according to the training data to obtain a target number of classification categories, the method further includes:
[0073] Preprocess the training data to obtain the eigenvalues;
[0074] The first object is classified according to the feature value to obtain a target number of classification categories.
[0075] Optionally, in the disclosed embodiment, the server may perform preprocessing work such as data cleaning, data deduplication, and data formatting on the training data to obtain feature values, and then classify the first object according to the feature values to obtain a target number of classification categories.
[0076] In addition, in order to avoid including the training data of historical drivers who may have cheated in the training data set, the server also eliminated the training data of drivers whose average daily service time exceeded 16 hours. The final training data set contained the training data of 67,866 historical drivers in total.
[0077] In the embodiment of the present disclosure, by performing data preprocessing on the training data and converting the training data into feature values, the quality and consistency of the training data can be ensured.
[0078] In some optional implementations, classifying the first object according to the training data to obtain a target number of classification categories includes:
[0079] According to the business service data of the first object, obtaining a first sub-dataset and a second sub-dataset from the training data set;
[0080] Obtaining a first number of classification categories according to the first training data in the first sub-dataset and a preset algorithm;
[0081] According to the first object dispatching service platform in the second sub-dataset, a second number of classification categories are obtained;
[0082] According to the first number of classification categories and the second number of classification categories, a target number of classification categories is obtained.
[0083] Optionally, in the disclosed embodiment, the first sub-dataset includes first training data, wherein the first training data is used to characterize the service capabilities of historical drivers, including the average daily average number of completed orders and the maximum daily average number of completed orders of historical drivers. The second sub-dataset includes service platform data of historical drivers’ dispatch, which is used to indicate the number of suppliers providing services to historical drivers. The preset algorithm is used to classify historical drivers according to the first training data, and the preset algorithm used in this embodiment is the K-Means Clustering Algorithm (KMA).
[0084] A clustering algorithm refers to dividing a data set into different classes according to a specific standard (such as distance) so that the similarity of data objects in the same class is as large as possible, and the difference of data objects in different classes is as large as possible. That is, after clustering, data of the same class are gathered together as much as possible, and data of different classes are separated as much as possible.
[0085] K-Means is a classic clustering algorithm that can divide data sets into different classes. The data points within each class have high similarity, while the data points between different classes have low similarity. It iteratively divides the data points into the class represented by the closest class center point, and then recalculates the class center point based on all the points in each class (taking the average value), and then repeats this process until the class center point changes very little or reaches the specified number of iterations.
[0086] First, the server classifies historical drivers based on their service capabilities:
[0087] Based on the silhouette coefficient and the historical average daily average number of completed orders and the maximum daily average number of completed orders of the drivers, the server obtains a first number of classification categories (eg, the first number in this embodiment is 5).
[0088] It should be noted that the silhouette coefficient is an indicator used to evaluate the clustering effect. It is an indicator used to describe the outline clarity of each category after clustering. It includes cohesion and separation. Cohesion is used to reflect the closeness between a sample point and the elements within the class, and separation is used to reflect the closeness between a sample point and the elements outside the class.
[0089] The server then categorizes the historical drivers based on the number of providers they provided service to:
[0090] The server obtains the second number of classification categories based on the service platform data of historical driver dispatch. For example, when the number of suppliers is 5, only the historical situation that the driver provides services for 1 supplier or 2 suppliers is considered, and the drivers can be divided into 15 categories.
[0091] Finally, the server multiplies the first number and the second number to obtain a target number (eg, the target number in this embodiment is 75), and then obtains a target number of classification categories based on the service capabilities of historical drivers and the number of service providers.
[0092] In the disclosed embodiment, by classifying the first object according to the first training data and the preset algorithm, a first number of classification categories are obtained, and the first object is classified in a first dimension. By classifying the first object according to the first object's vehicle dispatch service platform, a second number of classification categories are obtained, and the first object is classified in a second dimension. By obtaining a target number of classification categories according to the first number of classification categories and the second number of classification categories, the first object is classified in two different dimensions according to the differences of the first object in the training data.
[0093] In some optional implementations, before inputting the training data of the first object under each classification category into the corresponding initial model to obtain the initial prediction result, the method further includes:
[0094] Obtaining preset parameters, wherein the preset parameters are used to adjust the influence of the characteristic value;
[0095] Construct the corresponding initial model based on the feature values, preset parameters and current classification categories.
[0096] Optionally, in the embodiment of the present disclosure, the server constructs a corresponding logistic regression model according to the feature value X, the preset parameter β and the current classification category. The logistic regression model is defined as follows:
[0097] z=β0+β1X1+β2X2+...+β n X n
[0098] Here, z is the output of the logistic regression model.
[0099] In the embodiment of the present disclosure, by constructing a corresponding initial model according to the characteristic values, preset parameters and the current classification category, the input training data can be processed to obtain an initial prediction result.
[0100] In some optional implementations, inputting the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result includes:
[0101] Input the training data of the first object under each classification category into the corresponding initial model to obtain an output result;
[0102] An initial prediction result is obtained according to the output result and a preset function, wherein the preset function is used to convert the output result into a probability value, and the probability value is used to characterize the prediction result corresponding to the training data of the first object.
[0103] Optionally, in the embodiment of the present disclosure, a preset function is used to convert the output result z into a probability value. The preset function used in this embodiment is a Sigmoid function, which is defined as follows:
[0104]
[0105] Specifically, the server inputs the training data of historical drivers under each classification category into the corresponding logistic regression model to obtain the output result z, and then obtains the probability value σ(z) based on the output result z and the Sigmoid function to obtain the initial prediction result. The initial prediction result is defined as follows:
[0106] P(Y=1|X)=σ(z)
[0107] Among them, P(Y=1|X) represents the probability that the category label Y is equal to 1 given the feature X.
[0108] In the disclosed embodiment, by inputting the training data of the first object under each classification category into the corresponding initial model for training, the fitting ability of the initial model can be enhanced and the applicability of the initial model for the first object can be improved. By obtaining the initial prediction result according to the output result and the preset function, the output result of the initial model can be converted into the initial prediction result.
[0109] In some optional implementations, after adjusting preset parameters of a corresponding target number of initial models according to the initial prediction results and the sample labels to obtain a target number of target models, the method further includes:
[0110] Acquire target data of a second object for which the vehicle departure behavior is to be predicted;
[0111] Based on the target number of classification categories, determine the target classification category to which the target data belongs;
[0112] Input the target data into the target model corresponding to the target classification category to obtain the target prediction result;
[0113] obtaining a target behavior of a second object;
[0114] According to the target prediction result and the target behavior, the target parameters of the corresponding target model are adjusted to obtain the adjusted target model.
[0115] Optionally, in the embodiment of the present disclosure, the second object of the vehicle departure behavior to be predicted is the current driver in the actual application scenario, and the target behavior refers to the actual vehicle departure behavior of the current driver (including departure and non-departure).
[0116] Specifically, after obtaining the target number of target models, the server obtains the target data of the current driver, then preprocesses the target data, matches it with the target number of classification categories, determines the target classification category to which the target data belongs, and then inputs the target data into the target model corresponding to the target classification category to obtain the target prediction result, then obtains the actual driving behavior of the current driver, and uses the optimization algorithm to adjust the target parameters of the corresponding target model based on the target prediction result and the actual driving behavior to obtain the adjusted target model.
[0117] In the disclosed embodiment, by determining the target classification category to which the target data belongs based on the target number classification categories, the target data can be input into the target model corresponding to the target classification category according to the difference of the target data to obtain the target prediction result, thereby enhancing the applicability of the target model to the second object. By adjusting the target parameters of the corresponding target model according to the target prediction result and the target behavior, the target model can be optimized to obtain the adjusted target model.
[0118] In this embodiment, a device for determining an object vehicle departure prediction model is also provided, and the device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.
[0119] This embodiment provides a device for determining a vehicle departure prediction model of an object. Figure 2 As shown, including:
[0120] A first acquisition module 201 is used to acquire a first object vehicle dispatching training data set, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object vehicle dispatching behavior, and the sample label is used to characterize the first object vehicle dispatching specific behavior;
[0121] A first obtaining module 202 is used to classify the first object according to the training data to obtain a target number of classification categories;
[0122] The second obtaining module 203 is used to input the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result;
[0123] The third obtaining module 204 is used to adjust the preset parameters of the initial model according to the initial prediction results and sample labels to obtain the target model.
[0124] In the disclosed embodiment, a training data set for the first object's vehicle dispatch is obtained, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object's vehicle dispatch behavior, and the sample label is used to characterize the first object's specific vehicle dispatch behavior; the first object is classified according to the training data to obtain a target number of classification categories; the training data of the first object under each classification category is input into the corresponding initial model to obtain an initial prediction result; according to the initial prediction result and the sample label, the preset parameters of the initial model are adjusted to obtain a target model. Since the disclosed embodiment classifies the first object according to the difference in the training data, and inputs the training data of the first object under each classification category into the corresponding initial model, and adjusts the preset parameters of the initial model to obtain the target model, the fitting ability of the target model is enhanced, and the applicability of the target model to the first object is improved.
[0125] In some optional embodiments, the device further comprises:
[0126] The fourth obtaining module is used to preprocess the training data to obtain the characteristic value;
[0127] The fifth obtaining module is used to classify the first object according to the feature value to obtain a target number of classification categories.
[0128] In some optional implementations, the first obtaining module 202 includes:
[0129] An acquisition unit, configured to acquire a first sub-data set and a second sub-data set from a training data set according to the business service data of the first object;
[0130] A first obtaining unit, configured to obtain a first number of classification categories according to first training data in a first sub-data set and a preset algorithm;
[0131] A second obtaining unit is used to obtain a second number of classification categories according to the first object vehicle dispatch service platform in the second sub-dataset;
[0132] The third obtaining unit is used to obtain a target number of classification categories according to the first number of classification categories and the second number of classification categories.
[0133] In some optional embodiments, the device further comprises:
[0134] A second acquisition module is used to acquire preset parameters, wherein the preset parameters are used to adjust the influence of the characteristic value;
[0135] The construction module is used to construct the corresponding initial model according to the feature values, preset parameters and the current classification category.
[0136] In some optional implementations, the second obtaining module 203 includes:
[0137] a fourth obtaining unit, configured to input the training data of the first object under each classification category into the corresponding initial model to obtain an output result;
[0138] The fifth obtaining unit is used to obtain an initial prediction result according to the output result and a preset function, wherein the preset function is used to convert the output result into a probability value, and the probability value is used to characterize the prediction result corresponding to the training data of the first object.
[0139] In some optional embodiments, the device further comprises:
[0140] A third acquisition module is used to acquire target data of a second object for which the vehicle departure behavior is to be predicted;
[0141] A determination module, used to determine the target classification category to which the target data belongs based on the target number classification categories;
[0142] The sixth obtaining module is used to input the target data into the target model corresponding to the target classification category to obtain the target prediction result;
[0143] A fourth acquisition module, used to acquire a target behavior of the second object;
[0144] The seventh obtaining module is used to adjust the target parameters of the corresponding target model according to the target prediction result and the target behavior to obtain the adjusted target model.
[0145] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.
[0146] The device for determining the object departure prediction model in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.
[0147] The present disclosure also provides a computer device having the above Figure 2 The device for determining the object vehicle departure prediction model is shown.
[0148] See also Figure 3 , Figure 3 is a schematic diagram of a computer device provided by an optional embodiment of the present disclosure, such as Figure 3 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 3 A processor 10 is taken as an example.
[0149] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.
[0150] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.
[0151] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0152] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.
[0153] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.
[0154] The embodiments of the present disclosure also provide a computer-readable storage medium. The above-mentioned method according to the embodiments of the present disclosure can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium and downloaded through a network, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.
[0155] A part of the present disclosure may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present disclosure through the operation of the computer. Those skilled in the art should understand that the existence of computer program instructions in computer-readable media includes, but is not limited to, source files, executable files, installation package files, etc., and accordingly, the way in which computer program instructions are executed by a computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to the computer.
[0156] Although the embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present disclosure, and such modifications and variations are all within the scope defined by the appended claims.
Claims
1. A method for determining a vehicle dispatch prediction model, characterized in that: The method comprises: Acquire a first object vehicle dispatching training data set, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object vehicle dispatching behavior, and the sample label is used to characterize the first object vehicle dispatching specific behavior; Classify the first object according to the training data to obtain a target number of classification categories; Inputting the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result; According to the initial prediction result and the sample label, the preset parameters of the initial model are adjusted to obtain the target model.
2. The method according to claim 1, characterized in that: Before classifying the first object according to the training data to obtain a target number of classification categories, the method further includes: Performing data preprocessing on the training data to obtain characteristic values; The first object is classified according to the feature value to obtain the target number of classification categories.
3. The method according to claim 2, characterized in that The step of classifying the first object according to the training data to obtain a target number of classification categories includes: According to the business service data of the first object, acquiring a first sub-dataset and a second sub-dataset from the training data set; Obtaining a first number of classification categories according to the first training data in the first sub-dataset and a preset algorithm; According to the first object dispatching service platform in the second sub-dataset, a second number of classification categories are obtained; The target number of classification categories is obtained according to the first number of classification categories and the second number of classification categories.
4. The method according to claim 2, characterized in that: Before inputting the training data of the first object under each classification category into the corresponding initial model to obtain the initial prediction result, the method further includes: Acquire the preset parameter, wherein the preset parameter is used to adjust the influence of the characteristic value; The corresponding initial model is constructed according to the characteristic value, the preset parameters and the current classification category.
5. The method according to claim 4, characterized in that The step of inputting the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result includes: Inputting the training data of the first object under each classification category into the corresponding initial model to obtain an output result; The initial prediction result is obtained according to the output result and a preset function, wherein the preset function is used to convert the output result into a probability value, and the probability value is used to characterize the prediction result corresponding to the training data of the first object.
6. The method according to any one of claims 1 to 5, characterized in that: After adjusting the preset parameters of the target number of initial models according to the initial prediction results and the sample labels to obtain the target number of target models, the method further includes: Acquire target data of a second object for which the vehicle departure behavior is to be predicted; Based on the target number of classification categories, determining the target classification category to which the target data belongs; Inputting the target data into the target model corresponding to the target classification category to obtain a target prediction result; obtaining a target behavior of the second object; According to the target prediction result and the target behavior, the target parameters of the corresponding target model are adjusted to obtain an adjusted target model.
7. A device for determining a vehicle departure prediction model, characterized in that: The device comprises: A first acquisition module is used to acquire a first object vehicle dispatching training data set, wherein each training data in the training data set carries a sample label, wherein the training data is used to predict the first object vehicle dispatching behavior, and the sample label is used to characterize the first object vehicle dispatching specific behavior; A first obtaining module is used to classify the first object according to the training data to obtain a target number of classification categories; A second obtaining module is used to input the training data of the first object under each classification category into the corresponding initial model to obtain an initial prediction result; The third obtaining module is used to adjust the preset parameters of the initial model according to the initial prediction results and the sample labels to obtain the target model.
8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for determining the object vehicle departure prediction model according to any one of claims 1 to 6 by executing the computer instructions.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method for determining the object vehicle departure prediction model according to any one of claims 1 to 6.
10. A computer program product, characterized in that It includes computer instructions, and the computer instructions are used to enable a computer to execute the method for determining the object vehicle departure prediction model according to any one of claims 1 to 6.