CN107578107A - Model training method and device - Google Patents
Model training method and device Download PDFInfo
- Publication number
- CN107578107A CN107578107A CN201710672128.6A CN201710672128A CN107578107A CN 107578107 A CN107578107 A CN 107578107A CN 201710672128 A CN201710672128 A CN 201710672128A CN 107578107 A CN107578107 A CN 107578107A
- Authority
- CN
- China
- Prior art keywords
- functional unit
- group
- oriented
- unit group
- operation result
- 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
Landscapes
- Electrically Operated Instructional Devices (AREA)
Abstract
Specification discloses a kind of model training method and device, and applied to machine learning platform, the machine learning platform includes some functional units, the specified function that the functional unit is used in implementation model training flow, and the model training method includes:In response to building GC group command, oriented functional unit group is generated based on the execution sequence between selected multiple functional units and the multiple functional unit;It is received as multigroup parameter that the oriented functional unit group is set;In response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
Description
Technical field
This specification is related to machine learning techniques field, more particularly to a kind of model training method and device.
Background technology
Machine learning (Machine Learning, ML) is a multi-field cross discipline, be related to probability theory, statistics,
The multi-door subjects such as Approximation Theory, convextiry analysis, algorithm complex theory.Can be with by specific machine learning algorithm and training sample
Build simultaneously training machine learning model.The machine learning model completed for training, is applied under corresponding business scenario,
Can be with service guidance operation.
The content of the invention
In view of this, this specification provides a kind of model training method and device.
Specifically, this specification is achieved by the following technical solution:
A kind of model training method, applied to machine learning platform, the machine learning platform includes some functional units,
The specified function that the functional unit is used in implementation model training flow, the model training method include:
In response to building GC group command, based on the execution sequence between selected multiple functional units and the multiple functional unit
Generate oriented functional unit group;
It is received as multigroup parameter that the oriented functional unit group is set;
In response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
A kind of model training apparatus, applied to machine learning platform, the machine learning platform includes some functional units,
The specified function that the functional unit is used in implementation model training flow, the model training apparatus include:
Group unit is built, in response to building GC group command, based between selected multiple functional units and the multiple functional unit
Execution sequence generate oriented functional unit group;
Setting unit, it is received as multigroup parameter that the oriented functional unit group is set;
Running unit, in response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
A kind of model training apparatus, applied to machine learning platform, the machine learning platform includes some functional units,
The specified function that the functional unit is used in implementation model training flow, the model training apparatus include:
Processor;
For storing the memory of machine-executable instruction;
Wherein, referred to by reading and performing the machine corresponding with model training logic of the memory storage and can perform
Order, the processor are prompted to:
In response to building GC group command, based on the execution sequence between selected multiple functional units and the multiple functional unit
Generate oriented functional unit group;
It is received as multigroup parameter that the oriented functional unit group is set;
In response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
This specification can be based on selected multiple functional units and the multiple function group it can be seen from above description
Execution sequence between part generates oriented functional unit group, and multigroup circulatory function component groups based on setting can be with automatic cycle
Oriented functional unit group is run, the efficiency that model is established and trained can be effectively improved, save substantial amounts of human cost.
Brief description of the drawings
Fig. 1 is a kind of schematic flow sheet of model training method shown in the exemplary embodiment of this specification one.
Fig. 2 is a kind of schematic diagram of machine learning platform UI pages shown in the exemplary embodiment of this specification one.
Fig. 3 is the structural representation for a kind of model training apparatus shown in the exemplary embodiment of this specification one.
Fig. 4 is a kind of block diagram of model training apparatus shown in the exemplary embodiment of this specification one.
Embodiment
Here exemplary embodiment will be illustrated in detail, its example is illustrated in the accompanying drawings.Following description is related to
During accompanying drawing, unless otherwise indicated, the same numbers in different accompanying drawings represent same or analogous key element.Following exemplary embodiment
Described in embodiment do not represent all embodiments consistent with this specification.On the contrary, they are only and such as institute
The example of the consistent apparatus and method of some aspects be described in detail in attached claims, this specification.
It is only merely for the purpose of description specific embodiment in the term that this specification uses, and is not intended to be limiting this explanation
Book." one kind " of used singulative, " described " and "the" are also intended to bag in this specification and in the appended claims
Most forms are included, unless context clearly shows that other implications.It is also understood that term "and/or" used herein is
Refer to and any or all may be combined comprising the associated list items purpose of one or more.
It will be appreciated that though various information may be described using term first, second, third, etc. in this specification, but
These information should not necessarily be limited by these terms.These terms are only used for same type of information being distinguished from each other out.For example, do not taking off
In the case of this specification scope, the first information can also be referred to as the second information, and similarly, the second information can also be claimed
For the first information.Depending on linguistic context, word as used in this " if " can be construed to " ... when " or
" when ... " or " in response to determining ".
Fig. 1 is a kind of schematic flow sheet of model training method shown in the exemplary embodiment of this specification one.
Above-mentioned model training method can apply to machine learning platform, and the machine learning platform includes some function groups
Part, each functional unit can be used for one or more of implementation model training flow to specify function.For example, machine learning
Platform can include realizing the functional unit of data statistics function, realize the functional unit of text analyzing function, realize network point
Analyse functional unit of function etc..User can select corresponding functional unit according to the needs of modeling, and set each function group
Execution sequence between part.
Fig. 1 is refer to, above-mentioned model training method may comprise steps of:
Step 102, in response to building GC group command, based between selected multiple functional units and the multiple functional unit
Execution sequence generates oriented functional unit group.
The present embodiment machine learning platform can support the generation and maintenance of oriented functional unit group.For example, in machine
When learning platform models, oriented functional unit group can be generated based on conventional multiple functional units and its execution sequence and can
Preserve, when subsequently reusing, the oriented functional unit group can be copied, without being selected one by one again, improve modeling
Efficiency.
In the present embodiment, functional unit can be dragged to UI (User Interface, user interface), and set
After execution sequence, the foundation of oriented functional unit group is carried out.By taking the model shown in Fig. 2 as an example, the model includes 5 function groups
Part:
Data source 201, available for the acquisition data at data source;
Data prediction 202, pre-processed available for the data to getting, such as:Delete invalid data etc.;
The classification of logistic regression two 203, classification based training is carried out available for pretreated partial data;
Prediction 204, available for the quality based on pretreated another part data assessment logistic regression two classification 203;
Outcome evaluation 205, obtained quality results are assessed available for aggregation of forecasts 204, obtain the classification of logistic regression two 203
Training result.
In the present embodiment, the execution sequence of this 5 functional units is as indicated by the arrows of fig. 2.It is preferable to obtain quality
Model, the parameter of the classification of logistic regression two 203 can be constantly changed, can be determined relatively according to the training result of different parameters
Excellent parameter.And the parameter for constantly changing the classification of logistic regression two 203 is typically required repetition to determine the process of more excellent parameter
Execution logic returns two classification 203, prediction 204 and outcome evaluation 205.In view of this, two classification can be returned with logic-based
203rd, prediction 204 and the execution sequence between this 3 functional units of outcome evaluation 205 and this 3 functional units generate oriented
Functional unit group.
In one example, above-mentioned 3 functional units can be chosen in the UI shown in Fig. 2, such as:It can be irised out with mouse
Above-mentioned 3 functional units, refer to the dotted line frame shown in Fig. 2, and then click right, which is chosen, establishes oriented functional unit group.
In another example, functional unit group can also be established by other means, such as:Add in the page for build group
Add functional unit and set execution sequence etc., this specification is not particularly limited to this.
Step 104, it is received as multigroup parameter that the oriented functional unit group is set.
In the present embodiment, can be that built vertical oriented functional unit group sets multigroup parameter, every group of parameter can wrap
One or more parameter values are included, can specifically be determined by the quantity of parameter.Still by taking Fig. 2 as an example, the parameter is logistic regression two
The parameter of classification 203.
In one example, the set-up mode of multigroup parameter can be to set to sentences, such as:from k
=1to 100, step=1, the sentence represent that parameter k value is that step-length 1, i.e. this parameter set have from 1 to 100
100 groups, parameter k value is respectively 1,2 in this 100 groups of parameters ..., and 100.It is corresponding, every group of parameter can also be set
The condition of convergence, this is no longer going to repeat them for the present embodiment.
In another example, if not having any rule between multigroup parameter, every group of parameter can also be set respectively, this
Specification is not particularly limited to this.
Step 106, in response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
This specification can be based on selected multiple functional units and the multiple function group it can be seen from above description
Execution sequence between part generates oriented functional unit group, and multigroup circulatory function component groups based on setting can be with automatic cycle
Oriented functional unit group is run, the efficiency that model is established and trained can be effectively improved, save substantial amounts of human cost.
Based on the embodiment shown in Fig. 1, abovementioned steps 106 are running oriented function group according to multigroup parameter cyclic of setting
During part group, it can first judge whether the oriented functional unit group is first node in model training flow.Wherein, node can wrap
Include:Functional unit and functional unit group.
If it is determined that the oriented functional unit group is not the first node in model training flow, then can decide whether to preserve
There is the operation result of upstream node;If preserving the operation result of upstream node, can on the basis of the operation result root
The oriented functional unit group is run according to multigroup parameter cyclic;If not preserving the operation result of upstream node, can transport
The row upstream node, preserves the upstream node operation result, and according to described multigroup on the basis of the operation result
Parameter cyclic runs the oriented functional unit group, i.e., when carrying out model training first, first runs the oriented functional unit
The upstream node of group.
If it is determined that the oriented functional unit group is the first node in model training flow, then can be followed according to multigroup parameter
Oriented functional unit group described in inscription of loop.
In oriented functional unit group described in circular flow, i-th group of parameter can be used to run the oriented functional unit group,
And after operation, i is updated to i+1, wherein, i initial value can be 0, or 1.
Still by taking the model shown in Fig. 2 and functional unit group as an example, it is assumed that the ginseng of the classification of functional unit logistic regression two 203
Number only 1, multigroup parameter of setting are:From k=1to 100, step=1.After multigroup parameter is set, Ke Yiyun
Row functional unit group.
In this example, it may be determined that functional unit group is not the first node in model training flow, is then judged
The operation result of upstream node whether is preserved, that is, judges whether to preserve the operation result of data prediction 202.If do not protect
Deposit, then can be explained is to run first, then can be brought into operation from first node, and corresponding preservation operation result.Specifically, Ke Yixian
Service data source 201, then service data pretreatment 202.
Based on the operation result of data prediction 202, operation function component groups first, parameter k value is 1, is being run
After end, the operation result of functional unit group can be preserved, that is, preserves the result that outcome evaluation 205 exports.It is then possible to will ginseng
Number k is set to 2, and according to the operation result operation function component groups, and in operation result again of the data prediction 202 preserved
Afterwards, the result that outcome evaluation 205 exports is preserved.Then, parameter k can be set to 3, and according to the data prediction preserved
202 operation result operation function component groups, and after operation result again, the result that outcome evaluation 205 exports is preserved, successively
Analogize, until k is set into 100, untill circulation perform function component groups 100 times.
The present embodiment, can by setting functional unit group and its parameter in model training it can be seen from above description
To realize the circular flow of functional unit group, functional unit is pulled one by one without user to build model, without user gradually
Arrange parameter, the efficiency of model training is improved, save substantial amounts of human cost.
Based on above-described embodiment, after functional unit group runs, the multigroup ginseng of result queries instructions query can be passed through
Several operation results, such as:Machine learning platform may indicate that the operation result of multigroup parameter is illustrated in by front end browser or APP
In the same page, user's contrast is facilitated to check.Optionally, visual analyzing can also be generated according to the operation result of multigroup parameter
Report, facilitates customer analysis, this specification is not particularly limited to this.
Based on above-described embodiment, in the oriented functional unit group of circular flow, can also be shown with the animation effect of setting
Group's frame, with vivid show the process that circulation performs.In practical implementations, machine learning platform can indicate that front end is clear
Look at device or APP shows group's frame with the animation effect of above-mentioned setting.
Please continue to refer to Fig. 2, Fig. 2 dotted portion is group's frame of oriented functional unit group, in the oriented work(of circular flow
During energy component groups, can be circulated this group of frames clockwise, and more UI positions can be not take up by the way of the circulation of group's frame
On the premise of prompt user currently to circulate to perform oriented functional unit group.
It is, of course, also possible to prompted using other modes, such as:Loading icons, displaying progress bar etc. are shown, this
Specification is not particularly limited to this.
Corresponding with the embodiment of foregoing model training method, this specification additionally provides the implementation of model training apparatus
Example.
The embodiment of this specification model training apparatus can be applied on machine learning platform, the machine learning platform
Physics realization can be the equipment such as server.Device embodiment can be realized by software, can also pass through hardware or soft
The mode of combination of hardware is realized.It is by server where it as the device on a logical meaning exemplified by implemented in software
Processor corresponding computer program instructions in nonvolatile memory are read in internal memory what operation was formed.From hardware layer
For face, as shown in figure 3, being a kind of hardware structure diagram of this specification model training apparatus place server, except shown in Fig. 3
Processor, internal memory, outside network interface and nonvolatile memory, the usual root of server in embodiment where device
According to the actual functional capability of the server, other hardware can also be included, this is repeated no more.
Fig. 4 is a kind of block diagram of model training apparatus shown in the exemplary embodiment of this specification one.
Fig. 4 is refer to, the model training apparatus 300 can be applied on the server shown in earlier figures 3, included:
Build group unit 301, setting unit 302, running unit 303, storage unit 304 and query unit 305.
Wherein, group unit 301 is built, in response to building GC group command, based on selected multiple functional units and the multiple function
Execution sequence between component generates oriented functional unit group;
Setting unit 302, it is received as multigroup parameter that the oriented functional unit group is set;
Running unit 303, in response to operating instruction, the oriented functional unit is run according to multigroup parameter cyclic
Group.
Optionally, the running unit 303:
When the oriented functional unit group is not the first node in model training flow, judge whether to preserve upstream section
The operation result of point;
If preserving the operation result of upstream node, followed on the basis of the operation result according to multigroup parameter
Oriented functional unit group described in inscription of loop;
If not preserving the operation result of upstream node, the upstream node is run, preserves the upstream node operation
As a result, the oriented functional unit group and on the basis of the operation result is run according to multigroup parameter cyclic;
Wherein, the node includes:Functional unit and functional unit group.
Optionally, the running unit 303, the oriented functional unit group is run using i-th group of parameter, and run
Bi Hou, i is updated to i+1.
Storage unit 304, preserve the operation result of every group of parameter;
Query unit 305, instructed in response to result queries, the operation result of each group parameter is illustrated in the same page.
Optionally, it is described to build group unit 301, it is the oriented function also after the oriented functional unit group is generated
Component groups generate group's frame;
The running unit 303, also in oriented functional unit group described in circular flow, shown with the animation effect of setting
Group's frame.
The function of unit and the implementation process of effect specifically refer to and step are corresponded in the above method in said apparatus
Implementation process, it will not be repeated here.
For device embodiment, because it corresponds essentially to embodiment of the method, so related part is real referring to method
Apply the part explanation of example.Device embodiment described above is only schematical, wherein described be used as separating component
The unit of explanation can be or may not be physically separate, can be as the part that unit is shown or can also
It is not physical location, you can with positioned at a place, or can also be distributed on multiple NEs.Can be according to reality
Need to select some or all of module therein to realize the purpose of this specification scheme.Those of ordinary skill in the art are not
In the case of paying creative work, you can to understand and implement.
System, device, module or the unit that above-described embodiment illustrates, it can specifically be realized by computer chip or entity,
Or realized by the product with certain function.One kind typically realizes that equipment is computer, and the concrete form of computer can
To be personal computer, laptop computer, cell phone, camera phone, smart phone, personal digital assistant, media play
In device, navigation equipment, E-mail receiver/send equipment, game console, tablet PC, wearable device or these equipment
The combination of any several equipment.
Corresponding with the embodiment of foregoing model training method, this specification also provides another model training apparatus, should
For machine learning platform, the machine learning platform includes some functional units, and the functional unit is instructed for implementation model
Practice the specified function in flow, the model training apparatus includes:Processor and the storage for storing machine-executable instruction
Device.Wherein, processor and memory are generally connected with each other by internal bus.It is described to set in other possible implementations
It is standby to be also possible that external interface, can be communicated with other equipment or part.
In the present embodiment, can by reading and performing the machine corresponding with model training logic of the memory storage
Execute instruction, the processor are prompted to:
In response to building GC group command, based on the execution sequence between selected multiple functional units and the multiple functional unit
Generate oriented functional unit group;
It is received as multigroup parameter that the oriented functional unit group is set;
In response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
Optionally, when running the oriented functional unit group according to multigroup parameter cyclic, the processor is promoted
Make:
When the oriented functional unit group is not the first node in model training flow, judge whether to preserve upstream section
The operation result of point;
If preserving the operation result of upstream node, followed on the basis of the operation result according to multigroup parameter
Oriented functional unit group described in inscription of loop;
If not preserving the operation result of upstream node, the upstream node is run, preserves the upstream node operation
As a result, the oriented functional unit group and on the basis of the operation result is run according to multigroup parameter cyclic;
Wherein, the node includes:Functional unit and functional unit group.
Optionally, when running the oriented functional unit group according to multigroup parameter cyclic, the processor is promoted
Make:
The oriented functional unit group is run using i-th group of parameter, and after operation, i is updated to i+1.
Optionally, referred to by reading and performing the machine corresponding with model training logic of the memory storage and can perform
Order, the processor are also prompted to:
Preserve the operation result of every group of parameter;
Instructed in response to result queries, the operation result of each group parameter is illustrated in the same page.
Optionally, referred to by reading and performing the machine corresponding with model training logic of the memory storage and can perform
Order, the processor are also prompted to:
It is oriented functional unit all living creatures frame in groups after the oriented functional unit group is generated;
In oriented functional unit group described in circular flow, group's frame is shown with the animation effect of setting.
Corresponding with the embodiment of foregoing model training method, this specification also provides a kind of computer-readable storage medium
Matter, applied to machine learning platform, the machine learning platform includes some functional units, and the functional unit is used to realize mould
Specified function in type training flow, computer program is stored with the computer-readable recording medium, the program is processed
Device realizes following steps when performing:
In response to building GC group command, based on the execution sequence between selected multiple functional units and the multiple functional unit
Generate oriented functional unit group;
It is received as multigroup parameter that the oriented functional unit group is set;
In response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
Optionally, it is described that the oriented functional unit group is run according to multigroup parameter cyclic, including:
When the oriented functional unit group is not the first node in model training flow, judge whether to preserve upstream section
The operation result of point;
If preserving the operation result of upstream node, followed on the basis of the operation result according to multigroup parameter
Oriented functional unit group described in inscription of loop;
If not preserving the operation result of upstream node, the upstream node is run, preserves the upstream node operation
As a result, the oriented functional unit group and on the basis of the operation result is run according to multigroup parameter cyclic;
Wherein, the node includes:Functional unit and functional unit group.
Optionally, it is described that the oriented functional unit group is run according to multigroup parameter cyclic, including:
The oriented functional unit group is run using i-th group of parameter, and after operation, i is updated to i+1.
Optionally, in addition to:
Preserve the operation result of every group of parameter;
Instructed in response to result queries, the operation result of each group parameter is illustrated in the same page.
Optionally, in addition to:
It is oriented functional unit all living creatures frame in groups after the oriented functional unit group is generated;
In oriented functional unit group described in circular flow, group's frame is shown with the animation effect of setting.
It is above-mentioned that this specification specific embodiment is described.Other embodiments are in the scope of the appended claims
It is interior.In some cases, the action recorded in detail in the claims or step can be come according to different from the order in embodiment
Perform and still can realize desired result.In addition, the process described in the accompanying drawings not necessarily require show it is specific suitable
Sequence or consecutive order could realize desired result.In some embodiments, multitasking and parallel processing be also can
With or be probably favourable.
The preferred embodiment of this specification is the foregoing is only, it is all in this explanation not to limit this specification
Within the spirit and principle of book, any modification, equivalent substitution and improvements done etc., the model of this specification protection should be included in
Within enclosing.
Claims (11)
1. a kind of model training method, applied to machine learning platform, the machine learning platform includes some functional units, institute
The specified function that functional unit is used in implementation model training flow is stated, the model training method includes:
In response to building GC group command, based on the execution sequence generation between selected multiple functional units and the multiple functional unit
Oriented functional unit group;
It is received as multigroup parameter that the oriented functional unit group is set;
In response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
2. according to the method for claim 1, described run the oriented functional unit group according to multigroup parameter cyclic,
Including:
When the oriented functional unit group is not the first node in model training flow, judge whether to preserve upstream node
Operation result;
If preserving the operation result of upstream node, transported on the basis of the operation result according to multigroup parameter cyclic
The row oriented functional unit group;
If not preserving the operation result of upstream node, the upstream node is run, preserves the upstream node operation result,
And the oriented functional unit group is run according to multigroup parameter cyclic on the basis of the operation result;
Wherein, the node includes:Functional unit and functional unit group.
3. method according to claim 1 or 2, described to run the oriented functional unit according to multigroup parameter cyclic
Group, including:
The oriented functional unit group is run using i-th group of parameter, and after operation, i is updated to i+1.
4. the method according to claim 11, in addition to:
Preserve the operation result of every group of parameter;
Instructed in response to result queries, the operation result of each group parameter is illustrated in the same page.
5. the method according to claim 11, in addition to:
It is oriented functional unit all living creatures frame in groups after the oriented functional unit group is generated;
In oriented functional unit group described in circular flow, group's frame is shown with the animation effect of setting.
6. a kind of model training apparatus, applied to machine learning platform, the machine learning platform includes some functional units, institute
The specified function that functional unit is used in implementation model training flow is stated, the model training apparatus includes:
Group unit is built, in response to building GC group command, based on holding between selected multiple functional units and the multiple functional unit
Row is sequentially generated oriented functional unit group;
Setting unit, it is received as multigroup parameter that the oriented functional unit group is set;
Running unit, in response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
7. device according to claim 6, the running unit:
When the oriented functional unit group is not the first node in model training flow, judge whether to preserve upstream node
Operation result;
If preserving the operation result of upstream node, transported on the basis of the operation result according to multigroup parameter cyclic
The row oriented functional unit group;
If not preserving the operation result of upstream node, the upstream node is run, preserves the upstream node operation result,
And the oriented functional unit group is run according to multigroup parameter cyclic on the basis of the operation result;
Wherein, the node includes:Functional unit and functional unit group.
8. the device according to claim 6 or 7,
The running unit, the oriented functional unit group is run using i-th group of parameter, and after operation, i be updated to
i+1。
9. device according to claim 6, in addition to:
Storage unit, preserve the operation result of every group of parameter;
Query unit, instructed in response to result queries, the operation result of each group parameter is illustrated in the same page.
10. device according to claim 6, in addition to:
It is described to build group unit, it is oriented functional unit all living creatures side in groups also after the oriented functional unit group is generated
Frame;
The running unit, also in oriented functional unit group described in circular flow, the group is shown with the animation effect of setting
Frame.
11. a kind of model training apparatus, applied to machine learning platform, the machine learning platform includes some functional units,
The specified function that the functional unit is used in implementation model training flow, the model training apparatus include:
Processor;
For storing the memory of machine-executable instruction;
Wherein, by reading and performing the machine-executable instruction corresponding with model training logic of the memory storage, institute
Processor is stated to be prompted to:
In response to building GC group command, based on the execution sequence generation between selected multiple functional units and the multiple functional unit
Oriented functional unit group;
It is received as multigroup parameter that the oriented functional unit group is set;
In response to operating instruction, the oriented functional unit group is run according to multigroup parameter cyclic.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710672128.6A CN107578107A (en) | 2017-08-08 | 2017-08-08 | Model training method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710672128.6A CN107578107A (en) | 2017-08-08 | 2017-08-08 | Model training method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107578107A true CN107578107A (en) | 2018-01-12 |
Family
ID=61034375
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710672128.6A Pending CN107578107A (en) | 2017-08-08 | 2017-08-08 | Model training method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107578107A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110309462A (en) * | 2019-07-10 | 2019-10-08 | 腾讯科技(深圳)有限公司 | Method for exhibiting data and system |
CN110598868A (en) * | 2018-05-25 | 2019-12-20 | 腾讯科技(深圳)有限公司 | Machine learning model building method and device and related equipment |
CN113867608A (en) * | 2021-09-02 | 2021-12-31 | 浙江大华技术股份有限公司 | Method and device for establishing business processing model |
CN114546365A (en) * | 2022-04-27 | 2022-05-27 | 北京寄云鼎城科技有限公司 | Flow visualization modeling method, server, computer system and medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150317563A1 (en) * | 2014-05-01 | 2015-11-05 | International Business Machines Corporation | Predicting application performance on hardware accelerators |
CN105843873A (en) * | 2016-03-18 | 2016-08-10 | 北京物思创想科技有限公司 | System and method for managing data modeling |
CN106779088A (en) * | 2016-12-06 | 2017-05-31 | 北京物思创想科技有限公司 | Perform the method and system of machine learning flow |
-
2017
- 2017-08-08 CN CN201710672128.6A patent/CN107578107A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150317563A1 (en) * | 2014-05-01 | 2015-11-05 | International Business Machines Corporation | Predicting application performance on hardware accelerators |
CN105843873A (en) * | 2016-03-18 | 2016-08-10 | 北京物思创想科技有限公司 | System and method for managing data modeling |
CN106779088A (en) * | 2016-12-06 | 2017-05-31 | 北京物思创想科技有限公司 | Perform the method and system of machine learning flow |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110598868A (en) * | 2018-05-25 | 2019-12-20 | 腾讯科技(深圳)有限公司 | Machine learning model building method and device and related equipment |
CN110309462A (en) * | 2019-07-10 | 2019-10-08 | 腾讯科技(深圳)有限公司 | Method for exhibiting data and system |
CN113867608A (en) * | 2021-09-02 | 2021-12-31 | 浙江大华技术股份有限公司 | Method and device for establishing business processing model |
CN114546365A (en) * | 2022-04-27 | 2022-05-27 | 北京寄云鼎城科技有限公司 | Flow visualization modeling method, server, computer system and medium |
CN114546365B (en) * | 2022-04-27 | 2022-07-26 | 北京寄云鼎城科技有限公司 | Flow visualization modeling method, server, computer system and medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20230368024A1 (en) | Neural architecture search | |
Wu et al. | Machine learning at facebook: Understanding inference at the edge | |
CN110232183B (en) | Keyword extraction model training method, keyword extraction device and storage medium | |
CN104067282B (en) | Counter operation in state machine lattice | |
CN107578107A (en) | Model training method and device | |
US10984319B2 (en) | Neural architecture search | |
CN104704514B (en) | Modeling data generating process | |
CN107256156A (en) | Method and system for the detection in state machine | |
CN110945500A (en) | Key value memory network | |
US20180129933A1 (en) | Method and Apparatus for Processing Data Sequence | |
CN106874292A (en) | Topic processing method and processing device | |
CN109918499A (en) | A kind of file classification method, device, computer equipment and storage medium | |
CN111222039B (en) | Session recommendation method and system based on long-term and short-term interest combination | |
CN108196964A (en) | A kind of method and apparatus by Streams Module | |
CN108090042A (en) | For identifying the method and apparatus of text subject | |
CN110263328A (en) | A kind of disciplinary capability type mask method, device, storage medium and terminal device | |
CN108229286A (en) | Language model generates and application process, device, electronic equipment and storage medium | |
CN108881254A (en) | Intruding detection system neural network based | |
CN115344805A (en) | Material auditing method, computing equipment and storage medium | |
CN112200310B (en) | Intelligent processor, data processing method and storage medium | |
US12014276B2 (en) | Deterministic training of machine learning models | |
CN109284493A (en) | Random exam pool implementation method and electronic equipment in a kind of education cloud platform | |
KR20200061154A (en) | Method and apparatus of analyzing diagram containing visual and textual information | |
CN111177493B (en) | Data processing method, device, server and storage medium | |
CN107766416A (en) | Data analysing method, apparatus and system |
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 | ||
REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 1248359 Country of ref document: HK |
|
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180112 |
|
RJ01 | Rejection of invention patent application after publication |