CN117217362A - Business process prediction method, device, equipment and readable storage medium - Google Patents
Business process prediction method, device, equipment and readable storage medium Download PDFInfo
- Publication number
- CN117217362A CN117217362A CN202311048365.7A CN202311048365A CN117217362A CN 117217362 A CN117217362 A CN 117217362A CN 202311048365 A CN202311048365 A CN 202311048365A CN 117217362 A CN117217362 A CN 117217362A
- Authority
- CN
- China
- Prior art keywords
- business process
- process prediction
- feature vector
- event
- dimensional
- 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
- 238000000034 method Methods 0.000 title claims abstract description 153
- 230000008569 process Effects 0.000 title claims abstract description 105
- 239000013598 vector Substances 0.000 claims abstract description 66
- 238000013528 artificial neural network Methods 0.000 claims abstract description 38
- 238000005457 optimization Methods 0.000 claims abstract description 8
- 230000000694 effects Effects 0.000 claims description 13
- 238000013527 convolutional neural network Methods 0.000 claims description 10
- 230000015654 memory Effects 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 5
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 238000012549 training Methods 0.000 abstract description 9
- 238000010801 machine learning Methods 0.000 abstract description 2
- 238000004891 communication Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to the field of machine learning, and provides a business process prediction method, a business process prediction device, business process prediction equipment and a readable storage medium, wherein the business process prediction method comprises the following steps: constructing each business flow in the event log as an event track; converting the event track to obtain a high-dimensional sparse feature vector, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through a preset embedding layer; and inputting the low-dimensional dense feature vector into a business process prediction model to obtain a business process prediction result, wherein the business process prediction model is constructed based on a heterogeneous neural network. The application can obtain the characteristic representation of the task in different modes at the same time, so as to reduce the loss of information, improve the weighting algorithm of the optimization part of the multi-task learning model into self-adaptive dynamic weighting, and improve the training efficiency and the parallel prediction precision of a plurality of tasks.
Description
Technical Field
The present application relates to the field of machine learning technologies, and in particular, to a method, an apparatus, a device, and a readable storage medium for predicting a business process.
Background
The current common business process prediction method comprises the following steps: and carrying out business process prediction through the single-task learning model and carrying out business process prediction through the isomorphic multitask learning model.
The business process prediction method based on the single-task learning model can only learn one task at a time, generally splits the task into a plurality of independent subtasks when facing complex problems, integrates the results to obtain final output, and does not consider the relevance among the subtasks. Because correlation among tasks is not considered, common information among different tasks is lack of mining, and learning ability is reduced in the model training process.
The premise of the isomorphic multitask learning model-based business process prediction method is that training data of each task has the same characteristic representation, but a large number of problems are not satisfied, for example, text data and image data related to the tasks are provided, and under the condition of insufficient sample size, a classifier with higher reliability is obtained by utilizing shared information, so that the isomorphic multitask learning method cannot solve the problems. Meanwhile, due to the fact that correlation of real neutron tasks is quite complex, a common multi-task learning model often has negative migration, so that a phenomenon of 'teeterboard' appears in a model effect, namely performance of one task is usually improved by damaging performance of other tasks.
Disclosure of Invention
The application provides a business process prediction method, a business process prediction device, business process prediction equipment and a readable storage medium, which are used for solving the technical problem of low prediction precision in the existing business process prediction method.
The application provides a business process prediction method, which comprises the following steps:
constructing each business flow in the event log as an event track;
converting the event track to obtain a high-dimensional sparse feature vector, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through a preset embedding layer;
and inputting the low-dimensional dense feature vector into a business process prediction model to obtain a business process prediction result, wherein the business process prediction model is constructed based on a heterogeneous neural network.
According to the method for predicting the business processes provided by the application, the construction of each business process in the event log as the event track comprises the following steps:
extracting the event log as a plurality of columns of data comprising track identifiers, activity identifiers and timestamp identifiers;
and constructing the multiple columns of data into event tracks based on the preset task.
According to the business process prediction method provided by the application, the converting the event track to obtain the high-dimensional sparse feature vector comprises the following steps:
and converting each event track into a high-dimensional sparse feature vector based on a single-heat coding mode.
According to the business process prediction method provided by the application, the converting the high-dimensional sparse feature vector into the low-dimensional dense feature vector through the preset embedding layer comprises the following steps:
and inputting the high-dimensional sparse feature vector into a preset embedding layer, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through linear mapping.
According to the business process prediction method provided by the application, the business process prediction method further comprises the following steps:
and replacing the expert network in the preset isomorphic multitask learning model with a heterogeneous neural network, and constructing to obtain a business process prediction model.
According to the business process prediction method provided by the application, the heterogeneous neural network comprises a time circulation neural network, a convolution neural network and an attention neural network.
According to the business process prediction method provided by the application, the business process prediction method further comprises the following steps:
and when the business process prediction model is constructed, the adaptive dynamic weighting is used as a weighting part of model optimization.
The application also provides a business process prediction device, which comprises:
the event track construction module is used for constructing each business process in the event log into an event track;
the feature vector conversion module is used for converting the event track to obtain a high-dimensional sparse feature vector, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through a preset embedding layer;
and the business process prediction module is used for inputting the low-dimensional dense feature vector into a business process prediction model to obtain a business process prediction result, and the business process prediction model is constructed based on a heterogeneous neural network.
The application also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the business process prediction method according to any one of the above when executing the program.
The present application also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a business process prediction method as described in any of the above.
According to the business process prediction method, the business process prediction device, the business process prediction equipment and the readable storage medium, an expert network in a multi-task learning model is replaced by three basic component models, namely a convolutional neural network, a time-circulating neural network and an attention neural network, simultaneously, the characteristic representation of a task is obtained in different modes, and the weighting algorithm of an optimization part of the multi-task learning model is improved to be self-adaptive dynamic weighting. The method converts the execution of each business flow in the event log into a collection of event tracks, and reconstructs input data conforming to a prediction model; converting the reconstructed event track into feature codes, and converting the feature codes into vectors which are dense in low dimension and have correlations through an embedding layer; the expert network in the multi-task learning model is replaced by the three basic component models of the convolutional neural network, the time-circulating neural network and the attention neural network, so that the characteristic representation of the task is obtained in different modes at the same time, the loss of information is reduced, the weighting algorithm of the optimizing part of the multi-task learning model is improved to be self-adaptive dynamic weighting, the training efficiency is improved, the computing resources are saved, and the parallel prediction precision of a plurality of tasks is improved.
Drawings
In order to more clearly illustrate the application or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow diagram of a business process prediction method provided by the present application;
FIG. 2 is a second flow chart of the business flow prediction method according to the present application;
FIG. 3 is a schematic diagram of a business process prediction device according to the present application;
fig. 4 is a schematic structural diagram of an electronic device provided by the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Referring to fig. 1, the present application provides a business process prediction method, which includes:
step 100, constructing each business process in an event log as an event track;
the step 100 further includes:
step 110, extracting the event log into a plurality of columns of data comprising track identifiers, activity identifiers and timestamp identifiers;
and step 120, constructing the multiple columns of data into event tracks based on preset task requirements.
Specifically, as shown in fig. 2, the reconstruction process of the event log is as follows:
first, three columns of data including only a track identifier, an activity identifier, and a timestamp identifier are extracted from the event log L, where the representation may be l= { CaseID, active, time }, where CaseID represents a flow instance, active represents each activity performed in the flow instance, and time represents an execution time corresponding to each activity performed in the flow instance.
Then, an event log conforming to the input form of the prediction model is constructed according to task requirements. For example, if the result of the task flow predicts that an event track is { (active 1, time 1), (active 2, time 2), (active 3, time 3) }, the input data corresponding to the event track may be { [ (active 1, time 1) ], [ (active 1, time 1), (active 2, time 2) ], [ (active 1, time 1), (active 2, time 2), (active 3, time 3) ] } and the corresponding tag value is { active3, active3, comp }, and comp is a termination symbol, which indicates that the activity sequence has been executed.
Step 200, converting the event track to obtain a high-dimensional sparse feature vector, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through a preset embedding layer;
the step 200 further includes:
step 210, converting each event track into a high-dimensional sparse feature vector based on a single thermal encoding mode.
Step 220, inputting the high-dimensional sparse feature vector into a preset embedding layer, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through linear mapping.
Specifically, the process of encoding the event trace as a feature vector is as follows:
the sign of the event log is denoted as l= { σ 1 ，σ 2 ，σ 3 ，…，σ n And }, wherein σ i Represents the ith event trace, sigma i ＝<e 1 ，e 2 ，e 3 ，…，e n >(n＝|σ i I), each event trace is converted into a feature vector x= [ x ] 1 ，x 2 ，x 3 ，…，x p ]Wherein p represents the number of samples, x i Representing a set of feature sets.
Business processThe prediction task contains event activity and temporal features, each of which can be processed as a one-dimensional vector or two-dimensional matrix, depending on the needs of the prediction model. The active feature adopts an One-Hot coding mode (a single-Hot coding mode), namely the active feature is marked with 1 at the position with the activity, and the other positions are marked with 0; and taking the time interval as a time characteristic, and normalizing the time characteristic through the range of all values in the event log. Since the single thermal encoding has high-dimensional sparse features and no inherent association exists in the conversion process, the high-dimensional and sparse vector is converted into a low-dimensional and dense feature vector e= [ e ] through linear mapping by inputting the feature vector into an embedded layer 1 ，e 2 ，e 3 ，…，e n ]，e n ∈R d Where d represents the coding dimension of the feature vector.
And 300, inputting the low-dimensional dense feature vector into a business process prediction model to obtain a business process prediction result, wherein the business process prediction model is constructed based on a heterogeneous neural network.
Specifically, the construction process of the heterogeneous multitask learning prediction model (i.e., the business process prediction model in the present embodiment) is as follows:
the application provides a concept of heterogeneous private subnetworks, and aims at the characteristics of flow paths, the expert network in a homogeneous multitask learning model is replaced by three heterogeneous neural networks, namely CNN (Convolutional Neural Network ), LSTM (Long Short-Term Memory network) and Transformer model (a neural network), so that the characteristic representation of tasks is obtained in different modes at the same time, and the effect of losing effective information is reduced.
Specifically, the three heterogeneous neural networks are selected as main reasons of the sub-expert network in the heterogeneous multi-task learning model: (1) The LSTM has good modeling capability on the time sequence, in the prediction of the business flow, the past information state can be obtained through the LSTM network, the information is connected to be output as the final state, and the characteristics related to the task can be better obtained by considering the context relation; (2) The data contained in the business process instance is converted into space data, and the CNN network can better extract local characteristics by utilizing rolling and pooling operations; (3) The relationship between every two events is calculated in the same event track by the transducer network based on the attention mechanism, and in the process, the analysis between the two events is irrelevant to the distance and is only relevant to the event vector, so that the problem of long-term dependence does not exist.
In this embodiment, the expert network in the multi-task learning model is replaced with three basic component models, namely a convolutional neural network, a time-loop neural network and an attention neural network, and the feature representation of the task is obtained in different modes, and the weighting algorithm of the optimization part of the multi-task learning model is improved to be self-adaptive dynamic weighting. The method converts the execution of each business flow in the event log into a collection of event tracks, and reconstructs input data conforming to a prediction model; converting the reconstructed event track into feature codes, and converting the feature codes into vectors which are dense in low dimension and have correlations through an embedding layer; the expert network in the multi-task learning model is replaced by the three basic component models of the convolutional neural network, the time-circulating neural network and the attention neural network, so that the characteristic representation of the task is obtained in different modes at the same time, the loss of information is reduced, the weighting algorithm of the optimizing part of the multi-task learning model is improved to be self-adaptive dynamic weighting, the training efficiency is improved, the computing resources are saved, and the parallel prediction precision of a plurality of tasks is improved.
In one embodiment, the business process prediction method provided by the embodiment of the present application may further include:
and 400, replacing the expert network in the preset isomorphic multitask learning model with a heterogeneous neural network, and constructing to obtain a business process prediction model, wherein the heterogeneous neural network comprises a time circulation neural network, a convolution neural network and an attention neural network.
The business process prediction method provided by the embodiment of the application can further comprise the following steps:
and 500, when the business process prediction model is constructed, taking the adaptive dynamic weighting as a weighting part of model optimization.
Specifically, the application provides a heterogeneous multi-task learning model based on deep learning, which predicts a business process. The application provides a concept of heterogeneous private subnetworks, wherein an expert network in a multi-task learning model is replaced by three basic component models, namely a convolutional neural network, a time-circulating neural network and an attention neural network, and simultaneously, the characteristic representation of the task is acquired in different modes, so that the loss of information is reduced, and the parallel prediction precision of a plurality of tasks is improved.
The application improves the weighting algorithm of the optimization part of the multi-task learning model into self-adaptive dynamic weighting. Since the introduction of multiple task labels brings about multiple Loss, the Loss for multiple tasks was previously weighted in a specified ratio, in which case only one ratio effect can be verified at a time. Multiple experiments are needed to obtain the optimal ratio, the cost is high, and the optimal solution is not easy to find.
The self-adaptive dynamic weighting can automatically learn the Loss weighting proportion, the weighting proportion can be dynamically adjusted in the training process, and the training efficiency is improved, so that the computing resources are saved, the optimal state is easier to find, and the problems existing in the prior art are effectively relieved.
According to the embodiment, the Loss weighting proportion is automatically learned through self-adaptive dynamic weighting, the weighting proportion can be dynamically adjusted in the training process, and the training efficiency is improved, so that the computing resources are saved, the optimal state is easier to find, and the problems existing in the prior art are effectively relieved.
The business process prediction device provided by the application is described below, and the business process prediction device described below and the business process prediction method described above can be correspondingly referred to each other.
Referring to fig. 3, the present application further provides a business process prediction apparatus, including:
the event track construction module 301 is configured to construct each service flow in the event log as an event track;
the feature vector conversion module 302 is configured to convert the event track to obtain a high-dimensional sparse feature vector, and convert the high-dimensional sparse feature vector into a low-dimensional dense feature vector through a preset embedding layer;
and the business process prediction module 303 is configured to input the low-dimensional dense feature vector into a business process prediction model to obtain a business process prediction result, where the business process prediction model is constructed based on a heterogeneous neural network.
Optionally, the event track construction module includes:
an event log extracting unit for extracting the event log as a plurality of columns of data including a track identifier, an activity identifier and a timestamp identifier;
the event track construction unit is used for constructing the multiple columns of data into event tracks based on the preset task requirement.
Optionally, the feature vector conversion module includes:
the event track conversion unit is used for converting each event track into a high-dimensional sparse feature vector based on a single-thermal coding mode.
Optionally, the feature vector conversion module further includes:
and the feature vector conversion unit is used for inputting the high-dimensional sparse feature vector into a preset embedding layer, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through linear mapping.
Optionally, the business process prediction device further includes:
and the business process prediction model construction module is used for replacing the expert network in the preset isomorphic multitask learning model with a heterogeneous neural network to construct and obtain a business process prediction model.
Optionally, the heterogeneous neural network includes a time-loop neural network, a convolutional neural network, and an attention neural network.
Optionally, the business process prediction method further includes:
and the self-adaptive dynamic weighting module is used for taking the self-adaptive dynamic weighting as a weighting part of model optimization when the business process prediction model is constructed.
Fig. 4 illustrates a physical schematic diagram of an electronic device, as shown in fig. 4, which may include: processor 410, communication interface (Communications Interface) 420, memory 430 and communication bus 440, wherein processor 410, communication interface 420 and memory 430 communicate with each other via communication bus 440. Processor 410 may invoke logic instructions in memory 430 to perform the business process prediction method.
Further, the logic instructions in the memory 430 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method 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, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In yet another aspect, the present application also provides a non-transitory computer readable storage medium having stored thereon a computer program that, when executed by a processor, is implemented to perform the business process prediction method provided by the methods described above.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present application without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims (10)
1. A business process prediction method, comprising:
constructing each business flow in the event log as an event track;
converting the event track to obtain a high-dimensional sparse feature vector, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through a preset embedding layer;
and inputting the low-dimensional dense feature vector into a business process prediction model to obtain a business process prediction result, wherein the business process prediction model is constructed based on a heterogeneous neural network.
2. The business process prediction method according to claim 1, wherein the constructing each business process in the event log as an event track comprises:
extracting the event log as a plurality of columns of data comprising track identifiers, activity identifiers and timestamp identifiers;
and constructing the multiple columns of data into event tracks based on the preset task.
3. The business process prediction method according to claim 1, wherein the converting the event trace to obtain a high-dimensional sparse feature vector comprises:
and converting each event track into a high-dimensional sparse feature vector based on a single-heat coding mode.
4. The business process prediction method according to claim 3, wherein said converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector by a preset embedding layer comprises:
and inputting the high-dimensional sparse feature vector into a preset embedding layer, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through linear mapping.
5. The business process prediction method according to claim 1, wherein the business process prediction method further comprises:
and replacing the expert network in the preset isomorphic multitask learning model with a heterogeneous neural network, and constructing to obtain a business process prediction model.
6. The business process prediction method according to claim 5, wherein the heterogeneous neural network comprises a time-loop neural network, a convolutional neural network, and an attention neural network.
7. The business process prediction method according to claim 1, wherein the business process prediction method further comprises:
and when the business process prediction model is constructed, the adaptive dynamic weighting is used as a weighting part of model optimization.
8. A business process prediction apparatus, comprising:
the event track construction module is used for constructing each business process in the event log into an event track;
the feature vector conversion module is used for converting the event track to obtain a high-dimensional sparse feature vector, and converting the high-dimensional sparse feature vector into a low-dimensional dense feature vector through a preset embedding layer;
and the business process prediction module is used for inputting the low-dimensional dense feature vector into a business process prediction model to obtain a business process prediction result, and the business process prediction model is constructed based on a heterogeneous neural network.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the business process prediction method of any of claims 1 to 7 when the program is executed by the processor.
10. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor implements the business process prediction method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311048365.7A CN117217362A (en) | 2023-08-18 | 2023-08-18 | Business process prediction method, device, equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311048365.7A CN117217362A (en) | 2023-08-18 | 2023-08-18 | Business process prediction method, device, equipment and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117217362A true CN117217362A (en) | 2023-12-12 |
Family
ID=89041549
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311048365.7A Pending CN117217362A (en) | 2023-08-18 | 2023-08-18 | Business process prediction method, device, equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117217362A (en) |
-
2023
- 2023-08-18 CN CN202311048365.7A patent/CN117217362A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111898696B (en) | Pseudo tag and tag prediction model generation method, device, medium and equipment | |
| CN113222150B (en) | Quantum state transformation method and device | |
| CN112115744B (en) | Point cloud data processing method and device, computer storage medium and electronic equipment | |
| CN115223662A (en) | Data processing method, device, equipment and storage medium | |
| CN116827685B (en) | Dynamic defense strategy method of micro-service system based on deep reinforcement learning | |
| CN113222151A (en) | Quantum state transformation method and device | |
| CN117291259A (en) | Operator optimization method and device, electronic equipment and storage medium | |
| CN117217362A (en) | Business process prediction method, device, equipment and readable storage medium | |
| CN116342420A (en) | Method and system for enhancing mixed degraded image | |
| CN114445692B (en) | Image recognition model construction method and device, computer equipment and storage medium | |
| CN115965078A (en) | Classification prediction model training method, classification prediction method, device and storage medium | |
| CN115080745A (en) | Multi-scene text classification method, device, equipment and medium based on artificial intelligence | |
| CN110929118B (en) | Network data processing method, device, apparatus and medium | |
| CN113627514A (en) | Data processing method and device of knowledge graph, electronic equipment and storage medium | |
| CN114120447A (en) | Behavior recognition method and system based on prototype comparison learning and storage medium | |
| CN113537942A (en) | Method and system for increasing number of sample marks | |
| CN113760407A (en) | Information processing method, device, equipment and storage medium | |
| CN114363951B (en) | Inter-cell flow collaborative prediction method and device | |
| CN117388716B (en) | Battery pack fault diagnosis method, system and storage medium based on time sequence data | |
| CN116340765B (en) | Electricity larceny user prediction method and device, storage medium and electronic equipment | |
| Guo et al. | Error correcting output coding-based conditional random fields for web page prediction | |
| CN117219190A (en) | Training method, device, equipment, medium and program product of molecular generation model | |
| CN115130474A (en) | Named entity recognition method, device, equipment and storage medium | |
| CN117456268A (en) | Image multi-label classification method, device, equipment and storage medium | |
| CN117932157A (en) | Operation and maintenance intelligent detection method and system based on computer data processing |
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 |