CN109034367A - Neural network update method, device, computer equipment and readable storage medium storing program for executing - Google Patents
Neural network update method, device, computer equipment and readable storage medium storing program for executing Download PDFInfo
- Publication number
- CN109034367A CN109034367A CN201810960330.3A CN201810960330A CN109034367A CN 109034367 A CN109034367 A CN 109034367A CN 201810960330 A CN201810960330 A CN 201810960330A CN 109034367 A CN109034367 A CN 109034367A
- Authority
- CN
- China
- Prior art keywords
- network
- neural network
- current
- member set
- nervus opticus
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
The present invention relates to a kind of neural network update methods, the described method includes: obtaining current more new data, the peripheral sensory neuron set of initial neural network and nervus opticus member set, first nerves network is obtained in conjunction with the initial neural network and the peripheral sensory neuron set;Nervus opticus network is obtained according to the current more new data training first nerves network;Obtain the third nerve member set of the nervus opticus network;Target nerve network is obtained in conjunction with the initial neural network, the nervus opticus member set and the third nerve member set.The present invention program can be avoided neural network and overfitting occurs when constantly updating.
Description
Technical field
The present invention relates to field of computer technology, set more particularly to a kind of neural network update method, device, computer
Standby and readable storage medium storing program for executing.
Background technique
In machine learning and related fields, the computation model of artificial neural network is normally used for estimating or may rely on
A large amount of input and general unknown approximate function.Artificial neural network typically appears as " neuron " interconnected, it can
With from the calculated value of input, and can machine learning and pattern-recognition due to their self-adaptive property system.
When the data of training neural network are constantly updated, neural network can also update therewith, but not due to training data
Disconnected study, with the increase of the number of iterations, the noise being fitted in training data inevitably results in the phenomenon that generating over-fitting.
There is presently no the technologies that overfitting can be generated when can constantly update to avoid neural network.
Summary of the invention
The purpose of the present invention is to provide a kind of neural network update method, device computer equipment and readable storage mediums
Matter can eliminate overfitting caused by neural network continuous renewal process.
The purpose of the present invention is achieved through the following technical solutions:
A kind of neural network update method, which comprises
Current more new data, the peripheral sensory neuron set of initial neural network and nervus opticus member set are obtained, in conjunction with institute
It states initial neural network and the peripheral sensory neuron set obtains first nerves network;
Nervus opticus network is obtained according to the current more new data training first nerves network;
Obtain the third nerve member set of the nervus opticus network;
Target mind is obtained in conjunction with the initial neural network, the nervus opticus member set and the third nerve member set
Through network.
The peripheral sensory neuron set for obtaining current more new data, initial neural network in one of the embodiments,
With nervus opticus member set, first nerves network packet is obtained in conjunction with the initial neural network and the peripheral sensory neuron set
It includes:
Current more new data is obtained, the neuron for extracting preset ratio in initial neural network obtains first nerves metaset
It closes, using the neuronal ensemble not being extracted in the initial neural network as nervus opticus member set;
The current network parameter for obtaining the initial neural network, according to the current network parameter and the first nerves
Member set obtains the first nerves network.
Initial neural network, the nervus opticus member set described in the combination and described in one of the embodiments,
Third nerve member set obtains target nerve network
The peripheral sensory neuron set in the initial neural network is updated to the third nerve member set, is obtained
Target nerve network.
The network parameter of the target nerve network and the initial neural network is current in one of the embodiments,
Network parameter is identical.
The preset ratio is determined according to the data volume of the current more new data in one of the embodiments,.
A kind of neural network updating device, described device include:
First obtains module, for obtaining the peripheral sensory neuron set and second of current more new data, initial neural network
Neuronal ensemble obtains first nerves network in conjunction with the initial neural network and the peripheral sensory neuron set;
Second obtains module, for obtaining nervus opticus according to the current more new data training first nerves network
Network;
Third obtains module, for obtaining the third nerve member set of the nervus opticus network;
4th obtains module, in conjunction with the initial neural network, the nervus opticus member set and third mind
Gather through member and obtains target nerve network.
The first acquisition module includes: in one of the embodiments,
First acquisition unit extracts the neuron of preset ratio in initial neural network for obtaining current more new data
Peripheral sensory neuron set is obtained, using the neuronal ensemble not being extracted in the initial neural network as nervus opticus metaset
It closes;
Second acquisition unit, for obtaining the current network parameter of the initial neural network, according to the current network
Parameter and the peripheral sensory neuron set obtain the first nerves network.
In one of the embodiments, the described 4th obtain that module is specifically used for will be described in the initial neural network
Peripheral sensory neuron set is updated to the third nerve member set and obtains target nerve network.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the processing
Device realizes following steps when executing the computer program:
Current more new data, the peripheral sensory neuron set of initial neural network and nervus opticus member set are obtained, in conjunction with institute
It states initial neural network and the peripheral sensory neuron set obtains first nerves network;
Nervus opticus network is obtained according to the current more new data training first nerves network;
Obtain the third nerve member set of the nervus opticus network;
Target mind is obtained in conjunction with the initial neural network, the nervus opticus member set and the third nerve member set
Through network.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
Following steps are realized when row:
Current more new data, the peripheral sensory neuron set of initial neural network and nervus opticus member set are obtained, in conjunction with institute
It states initial neural network and the peripheral sensory neuron set obtains first nerves network;
Nervus opticus network is obtained according to the current more new data training first nerves network;
Obtain the third nerve member set of the nervus opticus network;
Target mind is obtained in conjunction with the initial neural network, the nervus opticus member set and the third nerve member set
Through network.
According to the scheme of aforementioned present invention, the peripheral sensory neuron set of current more new data, initial neural network is obtained
With nervus opticus member set, first nerves network, root are obtained in conjunction with the initial neural network and the peripheral sensory neuron set
Nervus opticus network is obtained according to the current more new data training first nerves network, is only updated in initial neural network
Peripheral sensory neuron set can all be updated to avoid entire initial neural network;Obtain the third mind of the nervus opticus network
Gather through member, obtains target in conjunction with the initial neural network, the nervus opticus member set and the third nerve member set
Neural network, the peripheral sensory neuron set in initial neural network are replaced by third nerve member set, nervus opticus member set
It remains unchanged, it is ensured that data that the parameter of neural network will not be thus continually updated change to changed beyond recognition, can effectively avoid
Overfitting.
Detailed description of the invention
Fig. 1 is the applied environment figure of neural network update method in one embodiment;
Fig. 2 is the flow diagram of neural network update method in one embodiment;
Fig. 3 is the flow diagram of neural network update method in another embodiment;
Fig. 4 is the structural block diagram of neural network updating device in one embodiment;
Fig. 5 is the structural block diagram of neural network updating device in another embodiment;
Fig. 6 is the internal structure chart of computer equipment in one embodiment;
Fig. 7 is the internal structure chart of computer equipment in another embodiment.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention more comprehensible, with reference to the accompanying drawings and embodiments, to this
Invention is described in further detail.It should be appreciated that the specific embodiments described herein are only used to explain the present invention,
And the scope of protection of the present invention is not limited.
Fig. 1 is the schematic diagram of internal structure of server in one embodiment.The server includes being connected by system bus
Processor, non-volatile memory medium, network interface, built-in storage, input unit.Wherein the non-volatile of the server is deposited
Storage media has operating system, further includes a kind of neural network updating device, the neural network updating device is for realizing a kind of mind
Through network update method.The processor supports the operation of entire server for providing calculating and control ability.In server
Built-in storage provides environment for the operation of the neural network updating device in non-volatile memory medium, and network interface is used for and it
His server or terminal communicate, and are such as sent to other servers when server response clicking operation can produce control command
Or terminal etc..Input unit is keyboard, mouse or touch screen etc..Specifically, server obtains current more new data, initial mind
Peripheral sensory neuron set and nervus opticus member set through network, in conjunction with the initial neural network and the first nerves metaset
It closes and obtains first nerves network;Server obtains nervus opticus according to the current more new data training first nerves network
Network;Server obtains the third nerve member set of the nervus opticus network;Server is in conjunction with the initial neural network, institute
It states nervus opticus member set and the third nerve member set obtains target nerve network.Wherein, server can be with independent
The server cluster of server either multiple servers composition is realized.It will be understood by those skilled in the art that the application mentions
The neural network update method of confession, can be applied not only in application environment shown in FIG. 1, can also apply but be not limited to various
In computer or server.
In one embodiment, as shown in Fig. 2, providing a kind of neural network update method, it is applied to Fig. 1 in this way
In server for be illustrated, comprising the following steps:
Step S101 obtains the peripheral sensory neuron set and nervus opticus metaset of current more new data, initial neural network
It closes, obtains first nerves network in conjunction with the initial neural network and the peripheral sensory neuron set;
For example, having 1000 neurons in initial neural network, that extracts 500 and is used as peripheral sensory neuron set, then
The other parameters of initial neural network are constant, and only 1000 original neurons become 500 present first nerves
Member set.
Step S102 obtains nervus opticus network according to the current more new data training first nerves network;
Step S103 obtains the third nerve member set of the nervus opticus network;
Specifically, third nerve member set namely train after peripheral sensory neuron set.
Step S104, in conjunction with the initial neural network, the nervus opticus member set and the third nerve member set
Obtain target nerve network.
In above-mentioned neural network update method, by the peripheral sensory neuron for obtaining current more new data, initial neural network
Set and nervus opticus member set obtain first nerves net in conjunction with the initial neural network and the peripheral sensory neuron set
Network obtains nervus opticus network according to the current more new data training first nerves network, only updates initial nerve net
Peripheral sensory neuron set in network can all be updated to avoid entire initial neural network;Obtain the nervus opticus network
Third nerve member set is obtained in conjunction with the initial neural network, the nervus opticus member set and the third nerve member set
Target nerve network is taken, the peripheral sensory neuron set in initial neural network is replaced by third nerve member set, nervus opticus
Member set remains unchanged, it is ensured that data that the parameter of neural network will not be thus continually updated change to changed beyond recognition, can have
Effect avoids overfitting.
For example, the neural network of existing one face for identification, initial neural network is carried out with multiple human face datas
What training obtained, currently more new data is the face of gorilla, if all neurons of initial neural network are all used into gorilla
Face data constantly training update, that neural network finally obtained possibly can not identify original face, and can only identify
Gorilla face in order to make the neural network finally obtained that can not only identify face, but also can identify the face of gorilla, therefore,
The partial nerve member in initial neural network is only randomly selected each time, with the face of gorilla by the with this partial nerve member
One neural network is trained.
In one of the embodiments, as shown in figure 3, the first of the current more new data of the acquisition, initial neural network
Neuronal ensemble and nervus opticus member set obtain the first mind in conjunction with the initial neural network and the peripheral sensory neuron set
Include: through network
Step S1011 obtains current more new data, and the neuron for extracting preset ratio in initial neural network obtains first
Neuronal ensemble, using the neuronal ensemble not being extracted in the initial neural network as nervus opticus member set;
Specifically, preset ratio can be determined according to the data volume of currently more new data, it, can when currently more new data is seldom
To randomly select the peripheral sensory neuron set of less ratio, when currently more amount of new data is very big, larger proportion can be extracted
Peripheral sensory neuron set is trained.
Step S1012 obtains the current network parameter of the initial neural network, according to the current network parameter and institute
It states peripheral sensory neuron set and obtains the first nerves network.
Specifically, current network parameter refers to that the input layer of neural network, output layer etc. are fixed not other than neuron
The parameter of change.
Initial neural network, the nervus opticus member set described in the combination and described in one of the embodiments,
Third nerve member set obtains target nerve network specifically:
The peripheral sensory neuron set in the initial neural network is updated to the third nerve member set to obtain
Target nerve network.
The network parameter of the target nerve network and the initial neural network is current in one of the embodiments,
Network parameter is identical.
That is, the current network parameter of initial neural network remains unchanged in entire training process, only change
The peripheral sensory neuron set extracted.
It should be understood that although each step in the flow chart of Fig. 2-3 is successively shown according to the instruction of arrow,
These steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly stating otherwise herein, these steps
Execution there is no stringent sequences to limit, these steps can execute in other order.Moreover, at least one in Fig. 2-3
Part steps may include that perhaps these sub-steps of multiple stages or stage are not necessarily in synchronization to multiple sub-steps
Completion is executed, but can be executed at different times, the execution sequence in these sub-steps or stage is also not necessarily successively
It carries out, but can be at least part of the sub-step or stage of other steps or other steps in turn or alternately
It executes.
In one of the embodiments, as shown in figure 4, providing a kind of neural network updating device, described device includes:
First obtains module 401, for obtaining the peripheral sensory neuron set and the of current more new data, initial neural network
Two neuronal ensembles obtain first nerves network in conjunction with the initial neural network and the peripheral sensory neuron set;
Second obtains module 402, for obtaining second according to the current more new data training first nerves network
Neural network;
Third obtains module 403, for obtaining the third nerve member set of the nervus opticus network;
4th obtains module 404, in conjunction with the initial neural network, the nervus opticus member set and the third
Neuronal ensemble obtains target nerve network.
In one of the embodiments, as shown in figure 5, the first acquisition module includes:
First acquisition unit 4011 extracts the mind of preset ratio in initial neural network for obtaining current more new data
Peripheral sensory neuron set is obtained through member, using the neuronal ensemble not being extracted in the initial neural network as nervus opticus member
Set;
Second acquisition unit 4012, for obtaining the current network parameter of the initial neural network, according to described current
Network parameter and the peripheral sensory neuron set obtain the first nerves network.
In one of the embodiments, it is described 4th obtain module 404 be specifically used for will be in the initial neural network
The peripheral sensory neuron set is updated to the third nerve member set and obtains target nerve network.
The network parameter of the target nerve network and the initial neural network is current in one of the embodiments,
Network parameter is identical.
The preset ratio is determined according to the data volume of the current more new data in one of the embodiments,.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 6.The computer equipment include the processor connected by device bus, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating device, computer program and data
Library.The built-in storage provides environment for the operation of operating device and computer program in non-volatile memory medium.The calculating
The database of machine equipment is used to store neural network and updates the data being related to.The network interface of the computer equipment is used for and outside
Terminal by network connection communication.To realize a kind of neural network update method when the computer program is executed by processor.
In one embodiment, a kind of computer equipment is provided, which can be terminal, internal structure
Figure can be as shown in Figure 7.The computer equipment includes processor, the memory, network interface, display connected by system bus
Screen and input unit.Wherein, the processor of the computer equipment is for providing calculating and control ability.The computer equipment is deposited
Reservoir includes non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system and computer journey
Sequence.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The network interface of machine equipment is used to communicate with external terminal by network connection.When the computer program is executed by processor with
Realize a kind of neural network update method.The display screen of the computer equipment can be liquid crystal display or electric ink is shown
Screen, the input unit of the computer equipment can be the touch layer covered on display screen, be also possible on computer equipment shell
Key, trace ball or the Trackpad of setting can also be external keyboard, Trackpad or mouse etc..
It will be understood by those skilled in the art that structure shown in Fig. 6-7, only part relevant to application scheme
The block diagram of structure, does not constitute the restriction for the computer equipment being applied thereon to application scheme, and specific computer is set
Standby may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment, including memory and processor are provided, is stored in memory
Computer program, the processor perform the steps of acquisition current more new data, initial neural network when executing computer program
Peripheral sensory neuron set and nervus opticus member set, obtained in conjunction with the initial neural network and the peripheral sensory neuron set
First nerves network;Nervus opticus network is obtained according to the current more new data training first nerves network;Obtain institute
State the third nerve member set of nervus opticus network;In conjunction with the initial neural network, the nervus opticus member set and described
Third nerve member set obtains target nerve network.
Processor executes acquisition current more new data, the initially mind when computer program in one of the embodiments,
Peripheral sensory neuron set and nervus opticus member set through network, in conjunction with the initial neural network and the first nerves metaset
Closing and obtaining first nerves network includes: to obtain current more new data, and the neuron for extracting preset ratio in initial neural network obtains
Peripheral sensory neuron set is obtained, using the neuronal ensemble not being extracted in the initial neural network as nervus opticus member set;
The current network parameter for obtaining the initial neural network is obtained according to the current network parameter and the peripheral sensory neuron set
Take the first nerves network.
Processor executes initial neural network, institute described in the combination when computer program in one of the embodiments,
State nervus opticus member set and the third nerve member set obtain target nerve network include: will be in the initial neural network
The peripheral sensory neuron set be updated to the third nerve member set and obtain target nerve network.
When processor executes computer program in one of the embodiments, the network parameter of the target nerve network and
The current network parameter of the initial neural network is identical.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program performs the steps of the first nerves metaset for obtaining current more new data, initial neural network when being executed by processor
Conjunction and nervus opticus member set obtain first nerves network in conjunction with the initial neural network and the peripheral sensory neuron set;
Nervus opticus network is obtained according to the current more new data training first nerves network;Obtain the nervus opticus network
Third nerve member set;In conjunction with the initial neural network, the nervus opticus member set and the third nerve member set
Obtain target nerve network.
The current more new data, initial of obtaining when computer program is executed by processor in one of the embodiments,
The peripheral sensory neuron set and nervus opticus member set of neural network, in conjunction with the initial neural network and the peripheral sensory neuron
It includes: to obtain current more new data that set, which obtains first nerves network, extracts the neuron of preset ratio in initial neural network
Peripheral sensory neuron set is obtained, using the neuronal ensemble not being extracted in the initial neural network as nervus opticus metaset
It closes;The current network parameter for obtaining the initial neural network, according to the current network parameter and the first nerves metaset
It closes and obtains the first nerves network.
Initial neural network described in combination when computer program is executed by processor in one of the embodiments,
It includes: by the initial neural network that the nervus opticus member set and the third nerve member set, which obtain target nerve network,
In the peripheral sensory neuron set be updated to the third nerve member set and obtain target nerve network.
The network parameter of target nerve network when computer program is executed by processor in one of the embodiments,
It is identical with the current network parameter of the initial neural network.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to protection of the invention
Range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (10)
1. a kind of neural network update method, which is characterized in that the described method includes:
Current more new data, the peripheral sensory neuron set of initial neural network and nervus opticus member set are obtained, in conjunction with described first
Beginning neural network and the peripheral sensory neuron set obtain first nerves network;
Nervus opticus network is obtained according to the current more new data training first nerves network;
Obtain the third nerve member set of the nervus opticus network;
Target nerve net is obtained in conjunction with the initial neural network, the nervus opticus member set and the third nerve member set
Network.
2. neural network update method according to claim 1, which is characterized in that the current more new data, just of obtaining
The peripheral sensory neuron set and nervus opticus member set of beginning neural network, in conjunction with the initial neural network and the first nerves
Member set obtains first nerves network
Current more new data is obtained, the neuron for extracting preset ratio in initial neural network obtains peripheral sensory neuron set, will
The neuronal ensemble not being extracted in the initial neural network is as nervus opticus member set;
The current network parameter for obtaining the initial neural network, according to the current network parameter and the first nerves metaset
It closes and obtains the first nerves network.
3. neural network update method according to claim 1, which is characterized in that initial nerve net described in the combination
Network, the nervus opticus member set and the third nerve member set obtain target nerve network
The peripheral sensory neuron set in the initial neural network is updated to the third nerve member set, obtains target
Neural network.
4. neural network update method according to claim 2, which is characterized in that the network of the target nerve network is joined
Number is identical with the current network parameter of the initial neural network.
5. neural network update method according to claim 2, which is characterized in that the preset ratio is according to described current
The data volume of more new data determines.
6. a kind of neural network updating device, which is characterized in that described device includes:
First obtains module, for obtaining current more new data, the peripheral sensory neuron set of initial neural network and nervus opticus
Member set obtains first nerves network in conjunction with the initial neural network and the peripheral sensory neuron set;
Second obtains module, for obtaining nervus opticus net according to the current more new data training first nerves network
Network;
Third obtains module, for obtaining the third nerve member set of the nervus opticus network;
4th obtains module, in conjunction with the initial neural network, the nervus opticus member set and third nerve member
Set obtains target nerve network.
7. neural network updating device according to claim 6, which is characterized in that described first, which obtains module, includes:
First acquisition unit, for obtaining current more new data, the neuron for extracting preset ratio in initial neural network is obtained
Peripheral sensory neuron set, using the neuronal ensemble not being extracted in the initial neural network as nervus opticus member set;
Second acquisition unit, for obtaining the current network parameter of the initial neural network, according to the current network parameter
The first nerves network is obtained with the peripheral sensory neuron set.
8. neural network updating device according to claim 6, which is characterized in that the 4th acquisition module is specifically used for
The peripheral sensory neuron set in the initial neural network is updated to the third nerve member set and obtains target nerve
Network.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 5 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 5 is realized when being executed by processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810960330.3A CN109034367A (en) | 2018-08-22 | 2018-08-22 | Neural network update method, device, computer equipment and readable storage medium storing program for executing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810960330.3A CN109034367A (en) | 2018-08-22 | 2018-08-22 | Neural network update method, device, computer equipment and readable storage medium storing program for executing |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109034367A true CN109034367A (en) | 2018-12-18 |
Family
ID=64626764
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810960330.3A Pending CN109034367A (en) | 2018-08-22 | 2018-08-22 | Neural network update method, device, computer equipment and readable storage medium storing program for executing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109034367A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109919304A (en) * | 2019-03-04 | 2019-06-21 | 腾讯科技(深圳)有限公司 | Neural network searching method, device, readable storage medium storing program for executing and computer equipment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160224892A1 (en) * | 2015-01-29 | 2016-08-04 | Panasonic Intellectual Property Management Co., Ltd. | Transfer learning apparatus, transfer learning system, transfer learning method, and recording medium |
CN106096727A (en) * | 2016-06-02 | 2016-11-09 | 腾讯科技(深圳)有限公司 | A kind of network model based on machine learning building method and device |
CN106845381A (en) * | 2017-01-16 | 2017-06-13 | 西北工业大学 | Sky based on binary channels convolutional neural networks composes united hyperspectral image classification method |
WO2017166155A1 (en) * | 2016-03-31 | 2017-10-05 | 富士通株式会社 | Method and device for training neural network model, and electronic device |
-
2018
- 2018-08-22 CN CN201810960330.3A patent/CN109034367A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160224892A1 (en) * | 2015-01-29 | 2016-08-04 | Panasonic Intellectual Property Management Co., Ltd. | Transfer learning apparatus, transfer learning system, transfer learning method, and recording medium |
WO2017166155A1 (en) * | 2016-03-31 | 2017-10-05 | 富士通株式会社 | Method and device for training neural network model, and electronic device |
CN106096727A (en) * | 2016-06-02 | 2016-11-09 | 腾讯科技(深圳)有限公司 | A kind of network model based on machine learning building method and device |
CN106845381A (en) * | 2017-01-16 | 2017-06-13 | 西北工业大学 | Sky based on binary channels convolutional neural networks composes united hyperspectral image classification method |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109919304A (en) * | 2019-03-04 | 2019-06-21 | 腾讯科技(深圳)有限公司 | Neural network searching method, device, readable storage medium storing program for executing and computer equipment |
CN109919304B (en) * | 2019-03-04 | 2021-07-02 | 腾讯科技(深圳)有限公司 | Image processing method, image processing device, readable storage medium and computer equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR102531593B1 (en) | Method and apparatus, device and readable storage medium for predicting power consumption | |
CN108491320A (en) | Exception analysis method, device, computer equipment and the storage medium of application program | |
CN108268645A (en) | Big data processing method and system | |
CN109767461A (en) | Medical image registration method, device, computer equipment and storage medium | |
US9672464B2 (en) | Method and apparatus for efficient implementation of common neuron models | |
CN113570048B (en) | Circuit simulation-based memristor array neural network construction and optimization method | |
CN106875004A (en) | Composite mode neuronal messages processing method and system | |
CN107908747A (en) | A kind of page generation method and device | |
CN108959271A (en) | Document creation method, device, computer equipment and readable storage medium storing program for executing are described | |
CN110377218B (en) | Data processing method, data processing device, computer equipment and storage medium | |
CN109063720A (en) | Handwritten word training sample acquisition methods, device, computer equipment and storage medium | |
CN109271917A (en) | Face identification method, device, computer equipment and readable storage medium storing program for executing | |
CN108769125A (en) | Using recommendation method, apparatus, storage medium and computer equipment | |
CN109767332A (en) | Asset Allocation strategy acquisition methods, device and computer equipment | |
CN113626610A (en) | Knowledge graph embedding method and device, computer equipment and storage medium | |
CN109034367A (en) | Neural network update method, device, computer equipment and readable storage medium storing program for executing | |
CN114707641A (en) | Training method, device, equipment and medium for neural network model of double-view diagram | |
CN112819687B (en) | Cross-domain image conversion method, device, computer equipment and storage medium based on unsupervised neural network | |
Caswell et al. | Loopy neural nets: Imitating feedback loops in the human brain | |
CN116864143A (en) | Infectious disease data analysis method and device based on artificial intelligence and related equipment | |
US20210042621A1 (en) | Method for operation of network model and related product | |
Gosmann | An Integrated Model of Contex, Short-Term, and Long-Term Memory | |
CN115730660A (en) | Link generation method based on causal decoupling and related device | |
CN115689649A (en) | Video advertisement putting method and device, computer equipment and storage medium | |
Sousa et al. | Building, simulating and visualizing large spiking neural networks with NeuralSyns |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181218 |