CN111275358A - Dispatch matching method, device, equipment and storage medium - Google Patents
Dispatch matching method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN111275358A CN111275358A CN202010118651.6A CN202010118651A CN111275358A CN 111275358 A CN111275358 A CN 111275358A CN 202010118651 A CN202010118651 A CN 202010118651A CN 111275358 A CN111275358 A CN 111275358A
- Authority
- CN
- China
- Prior art keywords
- task
- order
- matched
- execution
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The application provides a dispatching order matching method, a dispatching order matching device, dispatching order matching equipment and a storage medium, and relates to the technical field of model training. The method comprises the following steps: respectively calculating the success probability of the task corresponding to the order to be matched executed by each task execution object according to the difficulty information of the task corresponding to the order to be matched and the execution capacity information of at least one task execution object; determining a target object from at least one of the task execution objects according to at least one of the success probabilities; and distributing the task corresponding to the order to be matched to the target object. Compared with the prior art, the problem that the current worker executing the order task cannot complete the order task because the worker with corresponding capability cannot be matched for tasks with different difficulties is solved, so that the best effect cannot be achieved.
Description
Technical Field
The application relates to the technical field of model training, in particular to a method, a device, equipment and a storage medium for matching a dispatch.
Background
The work of a dispatch system generally includes: with the development of science and technology, the order dispatching system is applied more and more in life, such as: a taxi taking platform, a takeaway platform or a leg running platform, etc.
In the prior art, the dispatching system generally completes dispatching by actively grabbing orders by workers, or dispatches according to the distance between the orders and the workers, and selects the worker closest to the task of the current order as a target worker.
However, for tasks with different difficulty levels, the dispatching method cannot match workers with corresponding capabilities, so that the best effect cannot be achieved, and there may be a case that the worker currently executing the order task cannot complete the order task.
Disclosure of Invention
An object of the present application is to provide a method for matching orders, so as to solve the problem that in the prior art, for tasks with different difficulties, workers with corresponding capabilities cannot be matched, so that the best effect cannot be achieved, and a problem that a worker currently executing the order task cannot complete the order task may exist.
In order to achieve the above purpose, the technical solutions adopted in the embodiments of the present application are as follows:
in a first aspect, an embodiment of the present application provides a method for matching a dispatch, where the method includes:
respectively calculating the success probability of the task corresponding to the order to be matched executed by each task execution object according to the difficulty information of the task corresponding to the order to be matched and the execution capacity information of at least one task execution object;
determining a target object from at least one of the task execution objects according to at least one of the success probabilities;
and distributing the task corresponding to the order to be matched to the target object.
Optionally, the calculating, according to the difficulty information of the task corresponding to the to-be-matched order and the execution capability information of at least one task execution object, the success probability of the task corresponding to the to-be-matched order executed by each task execution object, respectively, includes:
respectively calculating relative difficulty information of the order to be matched relative to each task execution object according to the difficulty information and the execution capacity of at least one task execution object;
and determining the success probability according to the relative difficulty information.
Optionally, before the calculating the success probability of the task corresponding to the to-be-matched order executed by each task execution object according to the difficulty information of the task corresponding to the to-be-matched order and the execution capability information of at least one task execution object, the method further includes:
determining the execution capacity information of at least one task execution object by adopting a pre-trained first generation model according to the difficulty information of the task corresponding to the order to be matched;
the first generation model is a model obtained by training a plurality of first random data sets, and each first random data set comprises: the random task execution object executes an execution result of at least one random task, and has random task execution capacity, and each random task has a random task difficulty.
Optionally, before determining, according to the difficulty information of the task corresponding to the order to be matched, the execution capability information of at least one task execution object by using a pre-trained first generation model, the method further includes:
and processing the maximum likelihood function of the first generation model by adopting a regularization constraint algorithm according to preset average execution capacity information, so that the maximum likelihood function is converged to optimize the first generation model.
Optionally, before the calculating the success probability of the task corresponding to the to-be-matched order executed by each task execution object according to the difficulty information of the task corresponding to the to-be-matched order and the execution capability information of at least one task execution object, the method further includes:
determining the difficulty information by adopting a pre-trained second generation model according to the execution capacity information of at least one task execution object;
the second generation model is a model obtained by training a plurality of second random data sets, and each second random data set includes: at least one random task execution object executes an execution result of a random task, each random task execution object has a random task execution capability, and the random task has a random task difficulty.
Optionally, before determining the difficulty information by using a pre-trained second generative model according to the execution capability information of at least one task execution object, the method further includes:
and processing the maximum likelihood function of the second generation model by adopting a regularization constraint algorithm according to preset average task difficulty information, so that the maximum likelihood function is converged to optimize the second generation model.
In a second aspect, another embodiment of the present application provides an order matching apparatus, including: the device comprises a calculation module, a determination module and an allocation module, wherein:
the computing module is used for respectively computing the success probability of the task corresponding to the to-be-matched order executed by each task execution object according to the difficulty information of the task corresponding to the to-be-matched order and the execution capacity information of at least one task execution object;
the determining module is used for determining a target object from at least one task execution object according to at least one success probability;
and the distribution module is used for distributing the task corresponding to the order to be matched to the target object.
Optionally, the calculation module is further configured to calculate, according to the difficulty information and the execution capability of at least one task execution object, relative difficulty information of the order to be matched with respect to each task execution object respectively;
the determining module is further configured to determine the success probability according to the relative difficulty information.
Optionally, the determining module is further configured to determine, according to the difficulty information of the task corresponding to the order to be matched, execution capability information of at least one task execution object by using a pre-trained first generation model;
the first generation model is a model obtained by training a plurality of first random data sets, and each first random data set comprises: the random task execution object executes an execution result of at least one random task, and has random task execution capacity, and each random task has a random task difficulty.
Optionally, the apparatus further comprises: and the optimization module is used for processing the maximum likelihood function of the first generation model by adopting a regularization constraint algorithm according to preset average execution capacity information, so that the maximum likelihood function is converged to optimize the first generation model.
Optionally, the determining module is further configured to determine the difficulty information by using a pre-trained second generation model according to the execution capability information of at least one task execution object;
the second generation model is a model obtained by training a plurality of second random data sets, and each second random data set includes: at least one random task execution object executes an execution result of a random task, each random task execution object has a random task execution capability, and the random task has a random task difficulty.
Optionally, the optimization module is further configured to process the maximum likelihood function of the second generative model by using a regularization constraint algorithm according to preset average task difficulty information, so that the maximum likelihood function is converged to optimize the second generative model.
In a third aspect, another embodiment of the present application provides an order matching apparatus, including: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, when the order matching device runs, the processor and the storage medium communicate through the bus, and the processor executes the machine-readable instructions to perform the steps of the method according to any one of the first aspect.
In a fourth aspect, another embodiment of the present application provides a storage medium having a computer program stored thereon, where the computer program is executed by a processor to perform the steps of the method according to any one of the above first aspects.
By adopting the order dispatching matching method provided by the application, the probability of success of each task execution object in executing the order to be matched is respectively calculated according to the task difficulty information of the order to be matched and the execution capacity information of each task execution object, the target object is determined according to the success probability, and the order task to be matched is distributed to the target object. The order matching mode can fully consider the difficulty of the order to be matched and the capability of each task execution object during order dispatching, and distribute the order to be matched to the task performer with matched capability, thereby ensuring that the task can be performed smoothly and avoiding the condition that the task performer receiving the order can not complete the task of the order.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 is a schematic flow chart of a dispatch matching method according to an embodiment of the present application;
FIG. 2 is a schematic flow chart illustrating a method for matching a dispatch document according to another embodiment of the present application;
FIG. 3 is a schematic flow chart illustrating a method for matching a dispatch list according to another embodiment of the present application;
FIG. 4 is a schematic flow chart illustrating a method for matching a dispatch list according to another embodiment of the present application;
FIG. 5 is a schematic flow chart illustrating a method for matching a dispatch list according to another embodiment of the present application;
FIG. 6 is a schematic flow chart illustrating a method for matching a dispatch list according to another embodiment of the present application;
fig. 7 is a schematic structural diagram of a dispatch matching device according to an embodiment of the present application;
FIG. 8 is a schematic structural diagram of a dispatch matching device according to another embodiment of the present application;
fig. 9 is a schematic structural diagram of a dispatch matching device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments.
Fig. 1 is a schematic flow diagram of a matching method for a delivery, which is provided in an embodiment of the present application, and the method can be applied to any scene that needs to be matched for the delivery, such as a take-away delivery, a taxi taking delivery, a leg running delivery, and the like, a specific scene can be designed according to a user requirement, and is not limited to the above scene, and the method can be executed by a server or a service terminal, in an embodiment of the present application, an execution subject is taken as a service terminal for example: as shown in fig. 1, the method includes:
s101: and respectively calculating the success probability of the task corresponding to the to-be-matched order executed by each task execution object according to the difficulty information of the task corresponding to the to-be-matched order and the execution capacity information of at least one task execution object.
The execution capability information of each task execution object may include: expected performance capability information, and standard performance capability information.
In one embodiment of the present application, for an order to be matched, the difficulty of corresponding the order to the task may be x ∈ [0,1.0 ]. For a task execution object, the execution capability information shown when the task is completed can be y ∈ [0,1.0], wherein y obeys normal distribution y to N (μ, σ). Mu and sigma describe the execution capacity information of the task execution object, wherein mu represents the expected execution capacity information, and sigma represents the standard execution capacity information. Mu, sigma epsilon [0,1.0 ]. The success probability of the order to be matched executed by the task execution object corresponding to the task can be calculated based on y and x. When y ≧ x, the result of the task is a "pass" (pass), whose corresponding probability of success may be greater than or equal to a preset probability value, such as 0.5, for example; when y < x, the result of the task is "fail" (fail), and its corresponding success probability may be less than a preset probability value, such as 0.5, for example. In the actual application process, the preset probability value can be other values, different preset probability values can correspond to different system auditing standards, the higher the preset probability value is, the higher the success rate of the order matching is, the higher the success rate of the task execution is, and the smooth execution of the task is effectively ensured.
S102: a target object is determined from the at least one task execution object based on the at least one probability of success.
Optionally, when the target object is determined, the task execution object with the highest success probability may be selected according to the success probability, that is, the task execution object with the highest success probability is selected as the target object; or selecting task execution objects with the first ten successful probabilities, and then selecting an object closest to the order to be matched from the task execution objects with the first ten successful probabilities as a target object according to the distance between each task execution object and the order to be matched; or all task execution objects with the success probability of more than 80 percent are sent with the order tasks to be matched, and the order of each task execution object is taken, and the target object is the order task to be matched; the selection method of the specific target object may be designed according to the user's needs, and is not limited to the above embodiment.
S103: and distributing the task corresponding to the order to be matched to the target object.
Optionally, one or more target objects may be provided, and the present application is not limited in any way herein.
By adopting the order dispatching matching method provided by the application, the probability of success of each task execution object in executing the order to be matched is respectively calculated according to the task difficulty information of the order to be matched and the execution capacity information of each task execution object, the target object is determined according to the success probability, and the order task to be matched is distributed to the target object. The order matching mode can fully consider the difficulty of the order to be matched and the capability of each task execution object during order dispatching, and distribute the order to be matched to the task performer with matched capability, thereby ensuring that the task can be performed smoothly and avoiding the condition that the task performer receiving the order can not complete the task of the order.
Fig. 2 is a schematic flow chart of a method for matching a dispatch list according to another embodiment of the present application, and as shown in fig. 2, S101 may include:
s104: and respectively calculating the relative difficulty information of the order to be matched relative to each task execution object according to the difficulty information and the execution capacity of at least one task execution object.
Optionally, for example, in an embodiment of the present application, for a task corresponding to an order to be matched and a task execution object with execution capability information of y, F is then performed at this timey(x) And performing a cumulative distribution function of the corresponding execution capacity of the task execution object. I.e. Fy (x)
Wherein f isyIt is apparent that the probability of failure of the task execution object to execute the task for the pair of orders to be matched is Fy (x), i.e., Fy (x) is the relative difficulty information of the order to be matched with respect to each task execution object.
S105: and determining success probability according to the relative difficulty information.
In an embodiment of the present application, the success probability of determining the task corresponding to the executed to-be-matched order is:
P(y≧x│μ,σ)＝1-Fy(x)。
fig. 3 is a schematic flow chart of a method for matching a dispatch list according to another embodiment of the present application, as shown in fig. 3, before S101, the method further includes:
and S106, determining the execution capacity information of at least one task execution object by adopting a pre-trained first generation model according to the difficulty information of the task corresponding to the order to be matched.
The first generation model is a model obtained by training a plurality of first random data sets, and each first random data set comprises: the random task execution object is used for executing an execution result of at least one random task, the random task execution object has random task execution capacity, and each random task has a random task difficulty.
In an embodiment of the present application, assuming that the task execution object completes n tasks, the corresponding observation value is { x } x, where the task execution object completes n tasks, and the corresponding observation value is set as { x } x1,r1,x2,r2…xn,rnIn which xiIs task difficulty information, r, corresponding to each taskiAnd e (pass, fail) is the completion corresponding to each task. In an embodiment of the present application, the execution capability information μ, σ corresponding to the task execution object may be estimated according to the observation value, and the estimation may be performed by estimating μ, σ according to a maximum likelihood estimation problem, that is:
in one embodiment of the present application, the optimal μmay be iteratively calculated using a gradient ascent approach, i.e., based on
Calculating mu and sigma separately, where sμAnd sσThe learning rates of the gradient updates for mu and sigma respectively,
and
likelihood gradients corresponding to μ and σ, respectively; and calculating and updating the mu and the sigma according to the preset alternative training step number until the maximum likelihood function meets the preset convergence condition.
As can be seen from the above formula, the key to solving for μ and σ is to calculate
And
the following examples are respectively to
And
the calculation of (a) will be explained in detail.
First, calculate
When the temperature of the water is higher than the set temperature,
wherein, when r ≧ pass, P (x, r | mu, σ) ═ P (y ≧ x | mu, σ) ≧ 1-Fy(x) In that respect When r is fail, P (x, r | μ, σ) is P (y)<x│μ,σ)＝Fy(x) Thus, therefore, it is
It can be seen that for each pass or fail data point, the corresponding likelihood gradient is proportional to the inverse of the pass or fail probability, and inversely proportional to the pass or fail probability itself.
For pass data points, the greater the probability of passing, the less the effect on the gradient. For the data points of fail, the greater the probability of failure, the less the effect on the gradient. This means that the more the result is expected, the less the likelihood gradient is affected under the current parameters. Conversely, the occurrence of an unexpected result (such as a failure on a simple task or a pass on a difficult task) will produce a large likelihood gradient value indicating that the parameter needs to be updated.
For a data point of pass, its effect on the overall likelihood gradient is proportional to the derivative of the passing probability for that point on μ. Similarly, for a data point of fail, the effect on the total likelihood gradient is proportional to the derivative of the probability of failure for that point. This means that, under the current parameters, the smaller the derivative of the pass (fail) probability for the pass (fail) data point, the smaller the influence on the likelihood gradient. Otherwise large likelihood gradient values will result.
Due to the fact that
Is the inverse of the failure rate to mu, for any difficulty in the task to be performed, the failure rate decreases as mu increases in the task performer's performance, i.e., the failure rate decreases
If μ ═ x, then the success probability and the failure probability of the task to be executed at this time are equal, i.e., Fy(x) 0.5. When μ → ∞ Fy(x) → 0. When μ → - ∞ is reached, Fy(x) → 1.0. It can be seen that when μ ═ x,
maximum, the fastest change. When μ tends to ∞ or- ∞,
the change is slowest.
The likelihood function then solves for the gradient of μ as:
visible likelihood gradient
Relative difficulty with task Fy(x) The more difficult the task, the greater the likelihood gradient, the closer μ in the capability information of the executing task object is to the task difficulty x, and the greater the likelihood gradient.
Solving for
The method is similar to the above method, and the detailed description of the method is omitted here. Finally obtaining the final product
Alternatively, in one embodiment of the present application, in the case where the execution capability of the task execution object is unknown, the optimal μ and σ are learned by using an alternate update method, from the initial value μ0And σ0And starting to update the step [ mu ] of Q and then the step [ sigma ] of Q according to the preset alternative step numbers P and Q until the likelihood function converges.
By adopting the dispatch matching method provided by the embodiment, under the condition that the execution capacity of the task execution object is unknown, the execution capacity information of the task execution object can be estimated by adopting the preset first generation model according to the difficulty information of the task corresponding to the order to be matched, so that whether the task execution object has the capacity of executing the task to be matched is further judged, and even if the task execution object does not execute the task, the capacity of the task execution object can be estimated under the condition that the execution capacity is unknown, so that the corresponding order is matched.
Fig. 4 is a schematic flow chart of a method for matching a dispatch list according to another embodiment of the present application, as shown in fig. 4, before S106, the method further includes:
s107: and processing the maximum likelihood function of the first generation model by adopting a regularization constraint algorithm according to the preset average execution capacity information, so that the maximum likelihood function is converged to optimize the first generation model.
The preset average execution capacity information may include, for example, preset desired execution capacity information and preset standard execution capacity information.
Optionally, in one embodiment of the present application, μ*And σ*Regularization parameters corresponding to μ and σ, respectively, for μ, it is assumed that μ is a priori distributed over μ*In the vicinity, i.e. respectively according to
Optimizing the model; wherein λ isμAnd λσThe regularization strengths for μ and σ, respectively.
Optionally, in one embodiment of the present application, μ*And σ*0.5 and 5, respectively, in particular μ*And σ*The setting of (a) can be adjusted according to the user's needs, and is not limited to the above embodiment.
Due to the addition of the regularization constraint algorithm in the first generation model, regularization of the desired parameters is equivalent to introducing a prior distribution to the parameters (such constraints can be interpreted as a priori knowledge). The constraint has a guiding function, and the direction of gradient reduction meeting the constraint is prone to be selected when the error function is optimized, so that the final solution is prone to conform to the priori knowledge, and the calculation accuracy of mu and sigma is improved.
Fig. 5 is a schematic flow chart of a method for matching a dispatch list according to another embodiment of the present application, as shown in fig. 5, before S101, the method further includes:
s108: determining difficulty information by adopting a pre-trained second generation model according to the execution capacity information of at least one task execution object;
the second generation model is a model obtained by training a plurality of second random data sets, and each second random data set comprises: at least one random task execution object executes an execution result of a random task, each random task execution object has a random task execution capability, and the random task has a random task difficulty.
The specific algorithm for estimating the difficulty information according to the execution capacity information of the task execution object is similar to the above embodiment, but the difference is that for the first generation model, the known input information is the difficulty information of the task corresponding to the order to be matched, and for the second generation model, the known input information is the execution capacity information of the task execution object, and the rest of the similarities are referred to above, and are not repeated herein.
Fig. 6 is a schematic flow chart of a method for matching a dispatch list according to another embodiment of the present application, as shown in fig. 6, before S108, the method further includes:
s109: and processing the maximum likelihood function of the second generation model by adopting a regularization constraint algorithm according to preset average task difficulty information, so that the maximum likelihood function is converged to optimize the second generation model.
By adopting the dispatching matching method provided by the application, the executive capability information of each task executive object can be estimated according to the task difficulty information, and the task difficulty information can also be estimated according to the executive capability information of each task executive object; then, according to the task difficulty information of the order to be matched and the executive capability information of each task executive object, the probability of success of each task executive object in executing the order to be matched is respectively calculated, the target object is determined according to the probability of success, and the order task to be matched is distributed to the target object. The order matching mode can fully consider the difficulty of the order to be matched and the capability of each task execution object during order dispatching, and distribute the order to be matched to the task performer with matched capability, thereby ensuring that the task can be performed smoothly and avoiding the condition that the task performer receiving the order can not complete the task of the order.
Fig. 7 is a schematic structural diagram of a dispatch matching device according to an embodiment of the present application, and as shown in fig. 7, the device includes: a calculation module 201, a determination module 202 and an assignment module 203, wherein:
the calculating module 201 is configured to calculate, according to the difficulty information of the task corresponding to the to-be-matched order and the execution capability information of at least one task execution object, a success probability of the task corresponding to the to-be-matched order executed by each task execution object.
A determining module 202 configured to determine a target object from the at least one task execution object according to the at least one success probability.
The allocating module 203 is configured to allocate the task corresponding to the order to be matched to the target object.
Optionally, the calculating module 201 is further configured to calculate, according to the difficulty information and the execution capability of at least one task execution object, relative difficulty information of the to-be-matched order with respect to each task execution object.
The determining module 202 is further configured to determine a success probability according to the relative difficulty information.
Optionally, the determining module 202 is further configured to determine, according to the difficulty information of the task corresponding to the order to be matched, the execution capability information of at least one task execution object by using a pre-trained first generation model.
The first generation model is a model obtained by training a plurality of first random data sets, and each first random data set comprises: the random task execution object is used for executing an execution result of at least one random task, the random task execution object has random task execution capacity, and each random task has a random task difficulty.
Fig. 8 is a schematic structural diagram of a dispatch matching device according to another embodiment of the present application, and as shown in fig. 8, the device further includes: the optimization module 204 is configured to process the maximum likelihood function of the first generation model by using a regularization constraint algorithm according to preset average execution capacity information, so that the maximum likelihood function is converged, and the first generation model is optimized.
Optionally, the determining module 202 is further configured to determine the difficulty information by using a pre-trained second generative model according to the execution capability information of the at least one task execution object.
The second generation model is a model obtained by training a plurality of second random data sets, and each second random data set comprises: at least one random task execution object executes an execution result of a random task, each random task execution object has a random task execution capability, and the random task has a random task difficulty.
Optionally, the optimization module 204 is further configured to process the maximum likelihood function of the second generation model by using a regularization constraint algorithm according to preset average task difficulty information, so that the maximum likelihood function is converged, so as to optimize the second generation model.
The above-mentioned apparatus is used for executing the method provided by the foregoing embodiment, and the implementation principle and technical effect are similar, which are not described herein again.
These above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when one of the above modules is implemented in the form of a processing element scheduler code, the processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
Fig. 9 is a schematic structural diagram of a dispatch matching device according to an embodiment of the present application, where the dispatch matching device may be integrated in a server or a chip of the server, or alternatively, a service terminal or a chip of the service terminal.
As shown in fig. 9, the order matching apparatus includes: a processor 501, a storage medium 502, and a bus 503.
The processor 501 is used for storing a program, and the processor 501 calls the program stored in the storage medium 502 to perform the operations in any method embodiment corresponding to fig. 1-6. The specific implementation and technical effects are similar, and are not described herein again.
Optionally, the present application also provides a program product, such as a storage medium, on which a computer program is stored, including a program, which, when executed by a processor, performs embodiments corresponding to the above-described method.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to perform some steps of the methods according to the embodiments of the present application. And the aforementioned storage medium includes: a U disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
Claims (10)
1. A method of dispatch matching, the method comprising:
respectively calculating the success probability of the task corresponding to the order to be matched executed by each task execution object according to the difficulty information of the task corresponding to the order to be matched and the execution capacity information of at least one task execution object;
determining a target object from at least one of the task execution objects according to at least one of the success probabilities;
and distributing the task corresponding to the order to be matched to the target object.
2. The method of claim 1, wherein the calculating the success probability of the task corresponding to the to-be-matched order executed by each task execution object according to the difficulty information of the task corresponding to the to-be-matched order and the execution capability information of at least one task execution object respectively comprises:
respectively calculating relative difficulty information of the order to be matched relative to each task execution object according to the difficulty information and the execution capacity of at least one task execution object;
and determining the success probability according to the relative difficulty information.
3. The method according to claim 1 or 2, wherein before the calculating the success probability of the task corresponding to the to-be-matched order executed by each task execution object according to the difficulty information of the task corresponding to the to-be-matched order and the execution capability information of at least one task execution object, respectively, the method further comprises:
determining the execution capacity information of at least one task execution object by adopting a pre-trained first generation model according to the difficulty information of the task corresponding to the order to be matched;
the first generation model is a model obtained by training a plurality of first random data sets, and each first random data set comprises: the random task execution object executes an execution result of at least one random task, and has random task execution capacity, and each random task has a random task difficulty.
4. The method of claim 3, wherein before determining the execution capability information of at least one task execution object according to the difficulty information of the task corresponding to the order to be matched and by using a pre-trained first generation model, the method further comprises:
and processing the maximum likelihood function of the first generation model by adopting a regularization constraint algorithm according to preset average execution capacity information, so that the maximum likelihood function is converged to optimize the first generation model.
5. The method according to claim 1 or 2, wherein before the calculating the success probability of the task corresponding to the to-be-matched order executed by each task execution object according to the difficulty information of the task corresponding to the to-be-matched order and the execution capability information of at least one task execution object, respectively, the method further comprises:
determining the difficulty information by adopting a pre-trained second generation model according to the execution capacity information of at least one task execution object;
the second generation model is a model obtained by training a plurality of second random data sets, and each second random data set includes: at least one random task execution object executes an execution result of a random task, each random task execution object has a random task execution capability, and the random task has a random task difficulty.
6. The method of claim 5, wherein before determining the difficulty information using a pre-trained second generative model based on performance capability information of at least one of the task execution objects, the method further comprises:
and processing the maximum likelihood function of the second generation model by adopting a regularization constraint algorithm according to preset average task difficulty information, so that the maximum likelihood function is converged to optimize the second generation model.
7. An order matching apparatus, the apparatus comprising: the device comprises a calculation module, a determination module and an allocation module, wherein:
the computing module is used for respectively computing the success probability of the task corresponding to the to-be-matched order executed by each task execution object according to the difficulty information of the task corresponding to the to-be-matched order and the execution capacity information of at least one task execution object;
the determining module is used for determining a target object from at least one task execution object according to at least one success probability;
and the distribution module is used for distributing the task corresponding to the order to be matched to the target object.
8. The apparatus of claim 7, wherein the computing module is further configured to compute relative difficulty information of the order to be matched with respect to each task execution object according to the difficulty information and an execution capability of at least one task execution object;
the determining module is further configured to determine the success probability according to the relative difficulty information.
9. A dispatch matching device, comprising: a processor, a storage medium and a bus, the storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium communicating via the bus when the order matching device is running, the processor executing the machine-readable instructions to perform the method of any of claims 1-6.
10. A storage medium, characterized in that the storage medium has stored thereon a computer program which, when being executed by a processor, performs the method of any of the preceding claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010118651.6A CN111275358A (en) | 2020-02-25 | 2020-02-25 | Dispatch matching method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010118651.6A CN111275358A (en) | 2020-02-25 | 2020-02-25 | Dispatch matching method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111275358A true CN111275358A (en) | 2020-06-12 |
Family
ID=71003684
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010118651.6A Pending CN111275358A (en) | 2020-02-25 | 2020-02-25 | Dispatch matching method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111275358A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111738619A (en) * | 2020-07-06 | 2020-10-02 | 腾讯科技（深圳）有限公司 | Task scheduling method, device, equipment and storage medium |
| CN111882152A (en) * | 2020-06-18 | 2020-11-03 | 科大讯飞股份有限公司 | Task allocation method and related device |
| CN111967720A (en) * | 2020-07-22 | 2020-11-20 | 汉海信息技术(上海)有限公司 | Scheduling method and system for network appointment |
| CN113283628A (en) * | 2021-01-29 | 2021-08-20 | 汉海信息技术(上海)有限公司 | Information recommendation method and device, electronic equipment and readable storage medium |
| CN113869596A (en) * | 2021-10-12 | 2021-12-31 | 北京房江湖科技有限公司 | Task prediction processing method, device, product and medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108734432A (en) * | 2018-05-21 | 2018-11-02 | 北京顺丰同城科技有限公司 | Order allocation method and device |
| US20190066031A1 (en) * | 2017-08-30 | 2019-02-28 | 4D Technologies, LLC | Automatic Updating Interactive Query Answering and Feature Training Method and System |
| CN110163474A (en) * | 2019-04-12 | 2019-08-23 | 平安普惠企业管理有限公司 | A kind of method and apparatus of task distribution |
| CN110472910A (en) * | 2019-07-22 | 2019-11-19 | 北京三快在线科技有限公司 | Determine the method, apparatus and storage medium, electronic equipment of target dispatching task node |
-
2020
- 2020-02-25 CN CN202010118651.6A patent/CN111275358A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190066031A1 (en) * | 2017-08-30 | 2019-02-28 | 4D Technologies, LLC | Automatic Updating Interactive Query Answering and Feature Training Method and System |
| CN108734432A (en) * | 2018-05-21 | 2018-11-02 | 北京顺丰同城科技有限公司 | Order allocation method and device |
| CN110163474A (en) * | 2019-04-12 | 2019-08-23 | 平安普惠企业管理有限公司 | A kind of method and apparatus of task distribution |
| CN110472910A (en) * | 2019-07-22 | 2019-11-19 | 北京三快在线科技有限公司 | Determine the method, apparatus and storage medium, electronic equipment of target dispatching task node |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111882152A (en) * | 2020-06-18 | 2020-11-03 | 科大讯飞股份有限公司 | Task allocation method and related device |
| CN111738619A (en) * | 2020-07-06 | 2020-10-02 | 腾讯科技（深圳）有限公司 | Task scheduling method, device, equipment and storage medium |
| CN111738619B (en) * | 2020-07-06 | 2023-11-07 | 腾讯科技（深圳）有限公司 | Task scheduling method, device, equipment and storage medium |
| CN111967720A (en) * | 2020-07-22 | 2020-11-20 | 汉海信息技术(上海)有限公司 | Scheduling method and system for network appointment |
| CN111967720B (en) * | 2020-07-22 | 2024-03-08 | 汉海信息技术(上海)有限公司 | Scheduling method and system for network taxi taking |
| CN113283628A (en) * | 2021-01-29 | 2021-08-20 | 汉海信息技术(上海)有限公司 | Information recommendation method and device, electronic equipment and readable storage medium |
| CN113869596A (en) * | 2021-10-12 | 2021-12-31 | 北京房江湖科技有限公司 | Task prediction processing method, device, product and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111275358A (en) | Dispatch matching method, device, equipment and storage medium | |
| KR102170105B1 (en) | Method and apparatus for generating neural network structure, electronic device, storage medium | |
| CN113326126A (en) | Task processing method, task scheduling device and computer equipment | |
| CN110414569B (en) | Clustering implementation method and device | |
| JP7315007B2 (en) | LEARNING DEVICE, LEARNING METHOD AND LEARNING PROGRAM | |
| CN116257363B (en) | Resource scheduling method, device, equipment and storage medium | |
| CN111401642A (en) | Method, device and equipment for automatically adjusting predicted value and storage medium | |
| CN111860927B (en) | Model training method, service request processing method, device, equipment and medium | |
| JP2013105215A (en) | Recommendation information generation device, recommendation information generation method, and recommendation information generation program | |
| CN111510473B (en) | Access request processing method and device, electronic equipment and computer readable medium | |
| CN117193980A (en) | Task remaining duration calculation method and device | |
| CN117331668A (en) | Job scheduling method, device, equipment and storage medium | |
| US11514359B2 (en) | Distributed machine learning device, distributed machine learning method, and distributed machine learning recording medium | |
| CN111221827A (en) | Database table connection method and device based on graphic processor, computer equipment and storage medium | |
| CN112001570B (en) | Data processing method and device, electronic equipment and readable storage medium | |
| CN114912627A (en) | Recommendation model training method, system, computer device and storage medium | |
| CN114489966A (en) | Job scheduling method and device | |
| JP2022172503A (en) | Satellite observation planning system, satellite observation planning method and satellite observation planning program | |
| CN116755866B (en) | Resource scheduling method and device, electronic equipment and readable storage medium | |
| CN112650741A (en) | Abnormal data identification and correction method, system, equipment and readable storage medium | |
| CN113822455A (en) | Time prediction method, device, server and storage medium | |
| CN108958902B (en) | Graph calculation method and system | |
| CN113722014B (en) | Application program configuration method, device, equipment, storage medium and program product | |
| CN112561411B (en) | Method for calculating spare part safety stock number, storage equipment and terminal | |
| US11488056B2 (en) | Learning program, learning apparatus, and learning method |
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 |