CN114066076B - Network taxi appointment prediction method and device based on multiple tenants - Google Patents
Network taxi appointment prediction method and device based on multiple tenants Download PDFInfo
- Publication number
- CN114066076B CN114066076B CN202111387070.3A CN202111387070A CN114066076B CN 114066076 B CN114066076 B CN 114066076B CN 202111387070 A CN202111387070 A CN 202111387070A CN 114066076 B CN114066076 B CN 114066076B
- Authority
- CN
- China
- Prior art keywords
- data
- tenant
- historical
- target area
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000012937 correction Methods 0.000 claims description 44
- 230000015654 memory Effects 0.000 claims description 20
- 238000012549 training Methods 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 abstract description 9
- 230000002159 abnormal effect Effects 0.000 abstract description 6
- 230000000694 effects Effects 0.000 description 6
- 238000012795 verification Methods 0.000 description 6
- 238000011156 evaluation Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000010200 validation analysis Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000013486 operation strategy Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000009469 supplementation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0605—Supply or demand aggregation
-
- G06Q50/40—
Abstract
The invention provides a multi-tenant-based network car booking prediction method and device, wherein the multi-tenant-based network car booking prediction method comprises the following steps: acquiring data in a target area, wherein the data comprises tenant identification, order data, time data, weather data and grid index data of the tenant identification, the order data, the time data and the weather data in the target area; carrying out single-tenant supply and demand prediction based on tenant identification, order data, time data, weather data and the affiliated grid index data in the target area to obtain a network taxi appointment prediction result based on the single tenant; and integrating the network car booking prediction result based on the single tenant, and determining the network car booking prediction result based on the multiple tenants in the target area. The invention can respond and process quickly in time and ensure the stability of supply and demand relation in the area aiming at the situation of the sudden increase of the network taxi appointment orders caused by abnormal situations on the premise of ensuring timely and accurate prediction.
Description
Technical Field
The invention relates to the field of network car booking, in particular to a network car booking prediction method and device based on multiple tenants.
Background
The rapid development of the network car booking technology gradually replaces the traditional mode of 'roadside hiring a taxi' and the travel time and the destination are determined in advance on the network car booking platform according to actual conditions by the user, so that the convenience and the planning of the user in traveling are guaranteed. Most of the existing network car booking systems build a prediction model by collecting historical data, and real-time data is added to update and correct the prediction model, but the problem that abnormal conditions cannot be quickly processed still exists, if the network car booking orders are increased rapidly in rush hours, holidays or emergent weather and the like, timely and accurate prediction cannot be performed frequently, and the supply and demand relationship of network car booking in partial areas is unbalanced.
Disclosure of Invention
Therefore, the technical problem to be solved by the invention is to overcome the defect that the prediction model in the prior art cannot rapidly process abnormal conditions, so that the network car-booking prediction result cannot be timely and accurately reflected, thereby providing a network car-booking prediction method and device based on multiple tenants.
According to a first aspect, an embodiment of the present invention provides a network appointment prediction method based on multiple tenants, where the method includes:
acquiring data in a target area, wherein the data comprises tenant identification, order data, time data, weather data and grid index data of the tenant identification, the order data, the time data and the weather data in the target area;
performing single-tenant supply and demand prediction based on tenant identification, order data, time data, weather data and the affiliated grid index data in the target area to obtain a network taxi appointment prediction result based on the single tenant;
and integrating the network car booking prediction result based on the single tenant, and determining the network car booking prediction result based on the multi-tenant in the target area.
Optionally, the predicting the supply and demand of the single tenant based on the tenant identification, the order data, the time data, the weather data and the grid index data to which the tenant belongs in the target area to obtain the vehicle-booking prediction result of the network based on the single tenant comprises:
acquiring tenant identification, order data, time data and weather data in the target area and characteristic values of the grid index data;
and inputting the tenant identification, the order data, the time data, the weather data and the characteristic values of the grid index data in the target area into a prediction model to perform single-tenant supply and demand prediction, so as to obtain a network car-booking prediction result based on the single tenant.
Optionally, the prediction model is constructed by:
acquiring historical data in the target area, wherein the historical data comprises historical tenant identification, historical order data, historical time data, historical weather data and historical grid index data of the target area;
extracting the characteristics of the historical data in the target area to respectively obtain historical characteristic values corresponding to the historical data;
and constructing a prediction model based on the historical characteristic values.
Optionally, the method further comprises:
and training the prediction model based on the historical characteristic value corresponding to the historical data in the target area to obtain the final prediction model.
Optionally, the training the prediction model based on the historical feature value corresponding to the historical data in the target region to obtain a final prediction model includes:
inputting the sample data of the historical characteristic value into the prediction model to obtain a corresponding prediction value;
calculating a sample correction result based on the true value and the predicted value corresponding to the sample data;
based on the sample correction results, a final prediction model is determined.
Optionally, the sample correction result is calculated by the following formula:
wherein R is 2 adjusted represents the sample correction result, R 2 Denotes the corrected correlation coefficient, y i Indicating the true value corresponding to the ith sample data,
represents a predicted value corresponding to the ith sample data>
And representing the average value of the corresponding true values of all the sample data, wherein n is the number of samples, and p is the number of features.
Optionally, the determining a final prediction model based on the sample correction result includes:
judging whether the current sample correction result meets the preset correction result threshold value requirement or not;
when the current sample correction result meets the preset correction result threshold value requirement, determining the current prediction model as a final prediction model;
and when the current sample correction result does not meet the requirement of a preset correction result threshold, reselecting the sample data of the historical characteristic value and inputting the sample data into the prediction model to obtain a corresponding prediction value.
According to a second aspect, an embodiment of the present invention provides a multi-tenant-based network appointment prediction apparatus, including:
the acquisition module is used for acquiring data in a target area, wherein the data comprises tenant identification, order data, time data, weather data and grid index data of the target area;
the first processing module is used for carrying out single-tenant supply and demand prediction on the basis of tenant identification, order data, time data, weather data and grid index data of the target area to obtain a single-tenant-based network appointment prediction result;
and the second processing module is used for integrating the network car booking prediction result based on the single tenant and determining the network car booking prediction result based on the multi-tenant in the target area.
According to a third aspect, an embodiment of the present invention provides an electronic device, including:
a memory and a processor, the memory and the processor being communicatively coupled to each other, the memory having stored therein computer instructions, and the processor performing the method of the first aspect, or any one of the optional embodiments of the first aspect, by executing the computer instructions.
According to a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, which stores computer instructions for causing a computer to execute the method of the first aspect, or any one of the optional implementation manners of the first aspect.
The technical scheme of the invention has the following advantages:
according to the method and the device for predicting the network taxi appointment based on the multiple tenants, provided by the embodiment of the invention, data in a target area are obtained, wherein the data comprises tenant identification, order data, time data, weather data and grid index data of the target area; performing single-tenant supply and demand prediction based on tenant identification, order data, time data, weather data and the affiliated grid index data in the target area to obtain a network taxi appointment prediction result based on the single tenant; and integrating the network car booking prediction result based on the single tenant, and determining the network car booking prediction result based on the multi-tenant in the target area. Besides acquiring conventional data, the embodiment of the invention also acquires and processes time data and weather data in the target area, so that abnormal conditions can be predicted timely and accurately, single-tenant supply and demand prediction is performed to obtain a network appointment prediction result of a single tenant in the target area, and a network appointment prediction result based on multiple tenants in the target area is finally obtained by integrating network appointment prediction results of the single tenant.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a multi-tenant-based network taxi appointment prediction method according to an embodiment of the present invention;
fig. 2 is an application flowchart of a multi-tenant-based network taxi appointment prediction method according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a multi-tenant-based network appointment prediction device according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, it should be noted that the terms "first", "second", and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; the two elements may be directly connected or indirectly connected through an intermediate medium, or may be communicated with each other inside the two elements, or may be wirelessly connected or wired connected. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
In addition, the technical features involved in the different embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The embodiment of the invention provides a network car booking prediction method based on multiple tenants, and as shown in fig. 1, the network car booking prediction method based on the multiple tenants specifically comprises the following steps:
step S101: and acquiring data in the target area, wherein the data comprises tenant identification, order data, time data, weather data and affiliated grid index data in the target area. In practical applications, the time data includes week, hour, minute and whether holidays are saved, and the weather data includes temperature, humidity and specific weather conditions, such as sunny, fog, rain, etc., but the practical conditions are not limited to this, and the addition and supplementation of the data to ensure the data to be more reliable is also within the protection scope of the multi-tenant-based network appointment prediction method provided by the invention.
Specifically, in one embodiment, the grid index data is obtained through an open-source H3 geospatial index system, and the embodiment of the invention adopts 7-level resolution and tests, so that the supply and demand of the region, the modeling mobility and the optimization remarkable benefits in vehicle scheduling are better fed back on the premise of ensuring the data to be comprehensive. However, the practical situation is not limited to this, and it is also within the protection scope of the multi-tenant-based network appointment prediction method provided by the present invention to replace other spatial index systems or change the resolution for acquiring the corresponding data.
Step S102: and performing single-tenant supply and demand prediction based on tenant identification, order data, time data, weather data and the affiliated grid index data in the target area to obtain a network taxi appointment prediction result based on the single tenant.
Specifically, in an embodiment, the step S102 specifically includes the following steps:
step S201: and acquiring tenant identification, order data, time data, weather data and characteristic values of the grid index data in the target area.
Illustratively, the tenant identity, the order data, the time data, and the weather data, and the associated grid index data, and the corresponding feature values are shown in table 1 below. However, the actual situation is not limited to this, the various data and the feature values of the corresponding data in table 1 are only one representation provided by the embodiment of the present invention, for example, weather is represented by a numeral 4, and a method for selecting a numeral, a text or another representation for representing the various data and the feature values of the corresponding data is also within the protection scope of the multi-tenant-based network appointment prediction method provided by the present invention. The time interval for acquiring data in table 1 is 5 minutes, but the actual situation is not limited to this, and the change of the time interval to meet the accuracy of data acquisition is also within the protection scope of the multi-tenant-based network appointment prediction method provided by the present invention.
TABLE 1 characteristic variable correspondence table
Step S202: and inputting the tenant identification, the order data, the time data, the weather data and the characteristic values of the grid index data to a prediction model to predict the supply and demand of the single tenant, so as to obtain a network car-booking prediction result based on the single tenant. In practical application, due to the fact that the conditions of time and weather are considered, the internal evolution law of the data is summarized, and the result of the prediction model is more reliable.
Illustratively, in one embodiment, the characteristic values of the obtained data are shown in table 2, and the characteristic value of hour (hh) is 11, and the characteristic value of minute (mi) is 0, which indicates that the data is 11 hours and 00 minutes; the characteristic value of the grid index (order _ h3l 7) is 8740e3a1efffff, which indicates that the target area is a root; the characteristic value of the tenant identification (tent _ id) is 0, which indicates that the network appointment data of the tenant of the XX brand in the Chengdu area is viewed; a characteristic value of 2 for weather (weather) indicates that the weather condition on the same day is rain; the characteristic value of temperature (temperature) is 25, which means that the temperature is 25 ℃ on the same day; a characteristic value of humidity (humidity) of 56 indicates a humidity of 56% on the same day; a characteristic value of 0 for week (weekday) indicates that the day is monday; a characteristic value of 0 for holidays (holidays) indicates that the day is not a holiday. When an order is obtained, determining a selected time range through from _ x _ y, when x =0, indicating that data on the day is selected, and when x is a negative number, indicating that the data corresponds to data before several days; when y =0, data at the same time is shown, and when y is other values, data after calculating the interval time is shown, and since the value is taken every 5 minutes, the value of y is an integral multiple of 5. from _ 1_5 indicates that the order amount data 5 minutes after the same time point of yesterday, that is, the order data of 11 hours 00 to 05 minutes of yesterday, is selected, and it can be seen from the table that the order amount of 11 hours 00 to 05 minutes of yesterday is 3.
TABLE 2 comparison of eigenvalues with predicted results
Specifically, in an embodiment, the building of the prediction model in the step S202 specifically includes the following steps:
step S301: and acquiring historical data in the target area, wherein the historical data comprises historical tenant identification, historical order data, historical time data, historical weather data and historical grid index data of the target area. In practical application, the embodiment of the invention selects the target area, and then randomly selects a large amount of data in the database of the target area, thereby obtaining historical data.
Step S302: and extracting the characteristics of the historical data in the target area to respectively obtain historical characteristic values corresponding to the historical data. In practical applications, the method for obtaining the feature value of the historical data is the same as the method for obtaining the feature value in step S201, and is not described herein again.
Step S303: and constructing a prediction model based on the historical characteristic values. In practical application, the characteristic values of the historical data of the prediction model are constructed and include the characteristic values of the historical time data and the historical weather data, so that the prediction model based on the historical characteristic values is more sensitive to data change, can predict risks in advance and further has stronger exception handling capacity.
Specifically, in an embodiment, after obtaining the historical characteristic values in the target area, a training set, a testing set and a verification set are respectively generated, wherein the training set and the testing set are used for training a prediction model, so that the accuracy of the prediction model is ensured; the verification set is used for verifying the prediction effect of the prediction model, so that the prediction model is better adjusted, and the reliability of the prediction result is improved.
Specifically, in an embodiment, the step S303 further includes the following steps:
step S401: and training the prediction model based on the historical characteristic value corresponding to the historical data in the target area to obtain the final prediction model. In practical application, a Recurrent Neural Network (RNN) model is used in the embodiments of the present invention, and based on a training set and a test set, relevant parameters are adjusted until a loss function converges to generate an RNN model, and based on a validation set, an effect of the RNN model is validated, and an optimal RNN model, that is, a final prediction model, is obtained by repeating training and validation.
In practical application, the predictions of the network appointment vehicle, such as the time of getting on and off duty, weekends, holidays and the like, are closely related to the time sequence, and the embodiment of the invention adopts a specific model designed for the data, so that the time sequence information can be better processed, a better effect is achieved, and meanwhile, because the neural network model adopted by the embodiment of the invention does not need artificial characteristic extraction, the dependence on experience is avoided, and the network appointment vehicle has the advantages of strong generalization capability and strong expandability.
Specifically, in an embodiment, the step S401 further includes the following steps:
step S501: and inputting the sample data of the historical characteristic value into the prediction model to obtain a corresponding prediction value.
Step S502: and calculating a sample correction result based on the true value and the predicted value corresponding to the sample data. In practical application, a sample correction process is realized through the verification set, and a sample correction result is obtained. According to the embodiment of the invention, the correction result of the sample is calculated, so that the correction result is clearer and is convenient to check.
Specifically, in an embodiment, the formula for calculating the sample correction result in step S502 is as follows:
wherein R is 2 adjusted represents the sample correction result, R 2 Denotes the corrected correlation coefficient, y i Indicating the true value corresponding to the ith sample data,
represents a predicted value corresponding to the ith sample data>
And representing the average value of the corresponding true values of all the sample data, wherein n is the number of samples, and p is the number of features.
Step S503: based on the sample correction results, a final prediction model is determined. The prediction model is continuously corrected by correcting the sample, so that the reliability of the final prediction model is ensured, and the accuracy of the prediction result is finally ensured.
Specifically, in an embodiment, the step S503 specifically includes the following steps:
step S601: and judging whether the current sample correction result meets the preset correction result threshold value requirement or not. In practical application, the preset correction result threshold value can be adjusted according to the needs of the user, and the adjustment of the preset correction result threshold value is performed to meet the actual needs of the user, so that the method is also within the protection range of the multi-tenant-based network appointment prediction method.
Step S602: and when the current sample correction result meets the preset correction result threshold requirement, determining the current prediction model as the final prediction model.
Step S603: and when the current sample correction result does not meet the preset correction result threshold value requirement, reselecting sample data of the historical characteristic value and inputting the sample data into the prediction model to obtain a corresponding prediction value. In practical application, when the current sample correction result does not meet the preset correction result threshold requirement, the prediction model needs to be trained continuously, sample data is input into the prediction model again for training, and a corresponding predicted value is finally obtained, wherein the execution process is the same as that of step S401, and is not repeated here.
By executing the steps, the data acquisition time interval is more flexible, all scene conditions are covered, batch extension can be performed, and the prediction speed and the accuracy of the prediction result are greatly improved while the supply and demand prediction result of each tenant in the target area is acquired.
Step S103: and integrating the network car booking prediction result based on the single tenant, and determining the network car booking prediction result based on the multiple tenants in the target area. In practical application, the network car-booking prediction result based on multiple tenants in the target area can be obtained by summarizing the single-tenant network car-booking prediction result in the final prediction model. When a target area is determined, the net appointment prediction situations of a plurality of tenants in the target area are shown at the same time. Based on a multi-tenant mode, the real supply and demand of a target area can be predicted more accurately and more truly, scheduling tasks and releasing tasks are carried out more specifically, and finally correct guidance is given to scheduling tasks and vehicle optimization facing the future, empty driving of drivers is reduced, and the passenger taxi taking experience is improved.
By executing the steps, the network appointment prediction method based on the multi-tenant provided by the embodiment of the invention acquires data in the target area, wherein the data comprises tenant identification, order data, time data, weather data and grid index data of the target area; carrying out single-tenant supply and demand prediction based on tenant identification, order data, time data, weather data and grid index data of the tenant in the target area to obtain a single-tenant-based network taxi appointment prediction result; and integrating the network car booking prediction result based on the single tenant, and determining the network car booking prediction result based on the multiple tenants in the target area. Besides acquiring conventional data, the embodiment of the invention also acquires and processes time data and weather data in the target area, so that abnormal conditions can be predicted timely and accurately, single-tenant supply and demand prediction is performed to obtain a network appointment prediction result of a single tenant in the target area, and a network appointment prediction result based on multiple tenants in the target area is finally obtained by integrating network appointment prediction results of the single tenant.
The multi-tenant-based network appointment prediction method provided by the embodiment of the invention will be described in detail below with reference to specific application examples.
As shown in figure 2 of the drawings, in which,
1. randomly selecting a large amount of real historical data from an online database;
2. performing characteristic engineering processing on the data to generate a training set, a test set and a verification set;
3. repeating the processes of the step 1 and the step 2 to generate a plurality of groups of data;
4. using a Recurrent Neural Network (RNN):
1) Loading a training set and a test set to generate a data iterator for gradually calling an iteration model;
2) Initializing a model and related hyper-parameters, including a hidden layer, an output layer, gradient propagation, an additional gradient, a loss function, the number of training rounds, the size of each training data and the like;
3) Training is started until the loss function is converged, and an RNN reasoning model is generated, wherein the model can check the feature importance of each input dimension according to the influence/contribution degree of each dimension on the result Y in model training;
5. reading verification set data, inputting the data into a model of 4, and verifying the effect of 4 in the verification set;
6. and repeating the steps 4 and 5 to generate an optimal RNN model, wherein the evaluation indexes of the model mainly comprise:
1) Mean Squared Error, MSE (Mean Squared Error), is a measure reflecting the degree of difference between the estimated and estimated quantities, and is calculated as
Wherein m represents the number of samples, and when the prediction model is trained, m represents the number of samples in a training set; when performing predictive model validation, m represents the number of samples in the validation set, y i Indicating the true value corresponding to the ith sample data,
and indicating the predicted value corresponding to the ith sample data.
2) Correcting a decision coefficient (Adjusted R-Square), wherein the R-Square represents the variation degree which can be fitted by the model and accounts for the proportion of the variation degree of real data, the influence of the number of samples on the R-Square can be offset after correction, and the R-Square can be accurately evaluated when tens of millions of data are evaluated, and the calculation formula is
Wherein R is 2 adjusted represents the sample correction result, R 2 Denotes the corrected correlation coefficient, y i Indicating the true value corresponding to the ith sample data,
represents a predicted value corresponding to the ith sample data>
And representing the average value of the corresponding true values of all the sample data, wherein n is the number of samples, and p is the number of features.
7. Performing characteristic engineering processing on the real-time data;
8. inputting the real-time data in the step 7 into the model to obtain a future supply and demand prediction result in each region based on each tenant, wherein the on-line main evaluation indexes are as follows:
1) Giving out an evaluation result in real time by using the evaluation index in the step 6;
2) Because the method of the invention aims to reduce the idle running time of the driver, reduce the taxi taking response of passengers and improve the mobility of the city, the AB-Test can be carried out through the final income, and the significance of the invention is embodied. The method comprises the steps of providing two operation strategies A and B through AB-Test, namely based on a small sample, finally determining that the strategy A or the strategy B is more optimized through comparing operation results of the strategy A and the strategy B, further inputting the strategy A or the strategy B which can obtain a better operation result into a prediction model, continuously updating the prediction model through repeated iteration, finally obtaining the optimal prediction effect, and ensuring the stability of supply and demand relations in a region.
An embodiment of the present invention provides a multi-tenant based network car booking prediction apparatus, as shown in fig. 3, the multi-tenant based network car booking prediction apparatus includes:
the obtaining module 101 is configured to obtain data in the target area, where the data includes tenant identification, order data, time data, weather data, and grid index data to which the tenant identification, the order data, the time data, and the weather data belong in the target area. For details, refer to the related description of step S101 in the above method embodiment, and no further description is provided here.
The first processing module 102 is configured to perform single-tenant supply and demand prediction based on tenant identifiers, order data, time data, weather data, and grid index data to which the tenant identifiers, the order data, the time data, the weather data, and the grid index data belong in the target area, and obtain a single-tenant-based network car appointment prediction result. For details, refer to the related description of step S102 in the above method embodiment, and details are not repeated herein.
The second processing module 103 is configured to integrate the network car booking prediction result based on the single tenant, and determine a network car booking prediction result based on multiple tenants in the target area. For details, refer to the related description of step S103 in the above method embodiment, and no further description is provided here.
Through the cooperative cooperation of the components, the multi-tenant-based network appointment prediction device provided by the embodiment of the invention can better predict abnormal conditions timely and accurately by acquiring and processing time data and weather data in the target area, obtain the single-tenant network appointment prediction result in the target area by predicting the supply and demand of the single tenant, and finally obtain the multi-tenant-based network appointment prediction result in the target area by integrating the single-tenant network appointment prediction results.
An embodiment of the present invention provides an electronic device, as shown in fig. 4, the electronic device includes a processor 901 and a memory 902, and the memory 902 and the processor 901 are communicatively connected to each other, where the processor 901 and the memory 902 may be connected by a bus or in another manner, and fig. 4 takes the connection by the bus as an example.
The memory 902, which is a non-transitory computer readable storage medium, may be used for storing non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules corresponding to the methods of the embodiments of the present invention. The processor 901 executes various functional applications and data processing of the processor by executing non-transitory software programs, instructions and modules stored in the memory 902, that is, implements the methods in the above-described method embodiments.
The memory 902 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function; the storage data area may store data created by the processor 901, and the like. Further, the memory 902 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory 902 may optionally include memory located remotely from the processor 901, which may be connected to the processor 901 via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
One or more modules are stored in the memory 902, which when executed by the processor 901 perform the methods in the above-described method embodiments.
The specific details of the electronic device may be understood by referring to the corresponding related descriptions and effects in the above method embodiments, and are not described herein again.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, and the implemented program can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a Flash Memory (Flash Memory), a Hard Disk Drive (Hard Disk Drive, abbreviated as HDD), or a Solid State Drive (SSD); the storage medium may also comprise a combination of memories of the kind described above.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications therefrom are within the scope of the invention.
Claims (6)
1. A network taxi appointment prediction method based on multiple tenants is characterized by comprising the following steps:
acquiring data in a target area, wherein the data comprises tenant identification, order data, time data, weather data and grid index data of the tenant identification, the order data, the time data and the weather data in the target area;
performing single-tenant supply and demand prediction based on tenant identification, order data, time data, weather data and the affiliated grid index data in the target area to obtain a network taxi appointment prediction result based on the single tenant;
integrating the network car booking prediction result based on the single tenant, and determining a network car booking prediction result based on the multiple tenants in the target area;
the method for predicting the supply and demand of the single tenant based on the tenant identification, the order data, the time data, the weather data and the grid index data to which the tenant belongs in the target area to obtain the vehicle appointment prediction result of the network based on the single tenant comprises the following steps:
acquiring tenant identification, order data, time data and weather data in the target area and characteristic values of the grid index data;
inputting the tenant identification, the order data, the time data, the weather data and the characteristic values of the grid index data to a prediction model to perform single-tenant supply and demand prediction, and obtaining a network appointment prediction result based on the single tenant, wherein the prediction model is constructed in the following way:
acquiring historical data in the target area, wherein the historical data comprises historical tenant identification, historical order data, historical time data, historical weather data and historical grid index data of the target area;
extracting the characteristics of the historical data in the target area to respectively obtain historical characteristic values corresponding to the historical data;
constructing a prediction model based on the historical characteristic values;
training the prediction model based on a historical characteristic value corresponding to the historical data in the target area to obtain a final prediction model;
the training of the prediction model based on the historical characteristic value corresponding to the historical data in the target area to obtain a final prediction model comprises the following steps:
performing AB-Test on the sample data of the historical characteristic value to obtain operation results corresponding to the strategy A and the strategy B;
comparing the operation results of the strategy A and the strategy B to determine that the strategy A or the strategy B is more optimal;
inputting a strategy A or a strategy B which can obtain a better operation result into the prediction model for repeated iteration, and continuously updating the prediction model;
obtaining a corresponding predicted value based on the prediction model;
calculating a sample correction result based on the true value and the predicted value corresponding to the sample data;
based on the sample correction results, a final prediction model is determined.
2. The method of claim 1, wherein the sample correction is calculated by the formula:
wherein R is 2 adjusted represents the sample correction result, R 2 Denotes the corrected correlation coefficient, y i Represents the ith sampleThe true value to which the present data corresponds,
represents a predicted value corresponding to the ith sample data>
And representing the average value of the corresponding true values of all the sample data, wherein n is the number of samples, and p is the number of features.
3. The method of claim 2, wherein determining a final predictive model based on the sample correction results comprises:
judging whether the current sample correction result meets the preset correction result threshold value requirement or not;
when the current sample correction result meets the preset correction result threshold value requirement, determining the current prediction model as a final prediction model;
and when the current sample correction result does not meet the threshold requirement of the preset correction result, reselecting the sample data of the historical characteristic value to input into the prediction model to obtain a corresponding prediction value.
4. A network appointment prediction device based on multi-tenants is characterized by comprising:
the acquisition module is used for acquiring data in a target area, wherein the data comprises tenant identification, order data, time data, weather data and grid index data of the target area;
the first processing module is used for carrying out single-tenant supply and demand prediction on the basis of tenant identification, order data, time data, weather data and grid index data of the target area to obtain a single-tenant-based network appointment prediction result;
the second processing module is used for integrating the network car-booking prediction results based on the single tenant and determining the network car-booking prediction results based on the multiple tenants in the target area;
the method for predicting the supply and demand of the single tenant based on the tenant identification, the order data, the time data, the weather data and the grid index data to which the tenant belongs in the target area to obtain the vehicle appointment prediction result of the network based on the single tenant comprises the following steps:
acquiring tenant identification, order data, time data, weather data and characteristic values of the grid index data in the target area;
inputting the tenant identification, the order data, the time data, the weather data and the characteristic values of the grid index data in the target area into a prediction model to predict the supply and demand of the single tenant, and obtaining a network taxi-booking prediction result based on the single tenant, wherein the prediction model is constructed in the following mode:
acquiring historical data in the target area, wherein the historical data comprises historical tenant identification, historical order data, historical time data, historical weather data and historical grid index data of the target area;
extracting the characteristics of the historical data in the target area to respectively obtain historical characteristic values corresponding to the historical data;
constructing a prediction model based on the historical characteristic values;
training the prediction model based on a historical characteristic value corresponding to the historical data in the target area to obtain a final prediction model;
the training of the prediction model based on the historical characteristic value corresponding to the historical data in the target area to obtain a final prediction model comprises the following steps:
performing AB-Test on the sample data of the historical characteristic value to obtain operation results corresponding to the strategy A and the strategy B;
comparing the operation results of the strategy A and the strategy B to determine that the strategy A or the strategy B is more optimal;
inputting a strategy A or a strategy B which can obtain a better operation result into the prediction model for repeated iteration, and continuously updating the prediction model;
obtaining a corresponding predicted value based on the prediction model;
calculating a sample correction result based on the true value and the predicted value corresponding to the sample data;
based on the sample correction results, a final prediction model is determined.
5. An electronic device, comprising:
a memory and a processor communicatively coupled to each other, the memory having stored therein computer instructions, the processor executing the computer instructions to perform the method of any of claims 1-3.
6. A computer-readable storage medium having stored thereon computer instructions for causing a computer to thereby perform the method of any one of claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111387070.3A CN114066076B (en) | 2021-11-22 | 2021-11-22 | Network taxi appointment prediction method and device based on multiple tenants |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111387070.3A CN114066076B (en) | 2021-11-22 | 2021-11-22 | Network taxi appointment prediction method and device based on multiple tenants |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114066076A CN114066076A (en) | 2022-02-18 |
| CN114066076B true CN114066076B (en) | 2023-03-28 |
Family
ID=80278905
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111387070.3A Active CN114066076B (en) | 2021-11-22 | 2021-11-22 | Network taxi appointment prediction method and device based on multiple tenants |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114066076B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110458336A (en) * | 2019-07-23 | 2019-11-15 | 内蒙古工业大学 | A kind of net based on deep learning about vehicle supply and demand prediction method |
| CN111523693A (en) * | 2020-04-09 | 2020-08-11 | 北京白龙马云行科技有限公司 | Network car booking financial settlement method and device and network car booking SaaS platform |
| CN111753386A (en) * | 2019-03-11 | 2020-10-09 | 北京嘀嘀无限科技发展有限公司 | Data processing method and device |
| CN113538067A (en) * | 2021-08-17 | 2021-10-22 | 华侨大学 | Inter-city network taxi appointment demand prediction method and system based on machine learning |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104881706B (en) * | 2014-12-31 | 2018-05-25 | 天津弘源慧能科技有限公司 | A kind of power-system short-term load forecasting method based on big data technology |
| CN106952010A (en) * | 2017-01-04 | 2017-07-14 | 国家电网公司 | A kind of demand response system and method based on multi-tenant technology |
| CN108830504B (en) * | 2018-06-28 | 2021-09-21 | 清华大学 | Vehicle demand prediction method, system, server and computer storage medium |
| CN112561187B (en) * | 2020-12-22 | 2022-06-03 | 武汉理工大学 | Network taxi booking target order prediction method based on CNN-LSTM |
| CN113139726A (en) * | 2021-04-23 | 2021-07-20 | 北京白龙马云行科技有限公司 | Multi-tenant order dispatching method and network car booking system |
| CN112990610B (en) * | 2021-05-06 | 2021-08-20 | 北京工业大学 | Method for predicting taxi capacity demand of railway station based on multiple linear regression |
-
2021
- 2021-11-22 CN CN202111387070.3A patent/CN114066076B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111753386A (en) * | 2019-03-11 | 2020-10-09 | 北京嘀嘀无限科技发展有限公司 | Data processing method and device |
| CN110458336A (en) * | 2019-07-23 | 2019-11-15 | 内蒙古工业大学 | A kind of net based on deep learning about vehicle supply and demand prediction method |
| CN111523693A (en) * | 2020-04-09 | 2020-08-11 | 北京白龙马云行科技有限公司 | Network car booking financial settlement method and device and network car booking SaaS platform |
| CN113538067A (en) * | 2021-08-17 | 2021-10-22 | 华侨大学 | Inter-city network taxi appointment demand prediction method and system based on machine learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114066076A (en) | 2022-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107945507B (en) | Travel time prediction method and device | |
| US10936940B2 (en) | Arrangement and method for predicting road friction within a road network | |
| CN111445694B (en) | Festival and holiday traffic scheduling method and device based on traffic flow prediction | |
| US11507894B2 (en) | System and method for ride order dispatching | |
| US10331810B2 (en) | Method for determining a model of an output quantity of a technical system | |
| US20200125042A1 (en) | Control support device, apparatus control device, control support method, recording medium, learned model for causing computer to function, and method of generating learned model | |
| CN109637196A (en) | En-route sector traffic probability density prediction technique | |
| CN109344967B (en) | Intelligent electric meter life cycle prediction method based on artificial neural network | |
| CN109740818A (en) | A kind of probability density forecasting system applied to en-route sector traffic | |
| CN112381336A (en) | Flight delay duration prediction method and system | |
| CN111476389A (en) | Method and device for pre-estimating order receiving waiting time | |
| CN114274780A (en) | Vehicle driving mileage prediction method, device, equipment and storage medium | |
| CN112036598A (en) | Charging pile use information prediction method based on multi-information coupling | |
| CN113256472B (en) | Intelligent traffic control method and system and brain-like computer readable storage medium | |
| CN114066076B (en) | Network taxi appointment prediction method and device based on multiple tenants | |
| CN113222185A (en) | Analysis of vehicle drivelines in networked fleets | |
| CN114066079B (en) | Multi-tenant-based network appointment vehicle supply and demand difference prediction method and device | |
| CN115293366B (en) | Model training method, information prediction method, device, equipment and medium | |
| CN111507499A (en) | Construction method, test method, device and system of model for prediction | |
| CN116011995A (en) | Vehicle maintenance information pushing method and device, electronic equipment and storage medium | |
| CN114548463A (en) | Line information prediction method, line information prediction device, computer equipment and storage medium | |
| CN108985526B (en) | Transportation capacity prediction method and device, computer readable storage medium and terminal | |
| CN112308285A (en) | Information processing method, information processing device, electronic equipment and computer readable storage medium | |
| CN116662815B (en) | Training method of time prediction model and related equipment | |
| CN112967013B (en) | Method, system and electronic device for determining take-off time of pre-opened flight |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |