CN109919203A - A kind of data classification method and device based on Discrete Dynamic mechanism - Google Patents
A kind of data classification method and device based on Discrete Dynamic mechanism Download PDFInfo
- Publication number
- CN109919203A CN109919203A CN201910122983.9A CN201910122983A CN109919203A CN 109919203 A CN109919203 A CN 109919203A CN 201910122983 A CN201910122983 A CN 201910122983A CN 109919203 A CN109919203 A CN 109919203A
- Authority
- CN
- China
- Prior art keywords
- dynamic
- data
- module
- model
- layer
- 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
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of data classification method and device based on Discrete Dynamic mechanism, the device is based on a kind of data classification method based on Discrete Dynamic mechanism, comprising: receives set of data samples, carries out data prediction;Construct dynamic neural network model, every layer of the dynamic neural network model dynamic module established including several neurons, and the dynamic module in every layer is connect with a certain dynamic module of next layer, is fully-connected network between the feature that dynamic module extracts in the last layer and data category;Using the set of data samples training dynamic neural network model after standardization, data classification model is obtained;Data to be sorted are received, data classification is carried out according to data classification model.
Description
Technical field
The disclosure belongs to the technical field of data processing, be related to a kind of data classification method based on Discrete Dynamic mechanism and
Device.
Background technique
Only there is provided background technical informations relevant to the disclosure for the statement of this part, it is not necessary to so constitute first skill
Art.
Current era data have become a kind of important resource, are analyzed and processed to data with existing, therefrom find out rule
Rule, which carries out classification, to be one and significantly works.
Due to the limitation of hardware performance, deep learning modelling effect is not too much ideal, in recent years, graphics processor (GPU)
Appearance started the deep learning tide of a new wave, pushed the further development of artificial intelligence.In nearly all depth mould
In type, convolutional neural networks module (CNN) plays very crucial effect for the analysis and processing of big data.Currently, several
All deep learning models can all have different degrees of convolution module.Therefore, the research of convolutional neural networks is to depth
It is of great significance for habit.
In current convolutional neural networks model, if a kind of important selection mode of convolution kernel is selection random from image
Dry patch, is then based on selected patch and carries out feature extraction to figure, although many times achieving good effect
Fruit, but feature extraction is carried out using this method, the correlation between data fails sufficiently to be used, and model parameter is excessively huge
Greatly, usual number of parameters brings inconvenience to learning process all in 100,000,000 ranks.
Summary of the invention
For the deficiencies in the prior art, solve in data classification because data dependence is not in convolutional neural networks
Can obtain sufficiently using and model parameter excessively lead to that data classification performance is undesirable, the slow problem low with accuracy of speed, this public affairs
The one or more embodiments opened provide a kind of data classification method and device based on Discrete Dynamic mechanism, more fully sharp
It is suitable for text data, figure so that the internal information of more preferable mining data carries out data classification with the dependence between information
As the classification of the data such as data and audio.
According to the one aspect of one or more other embodiments of the present disclosure, a kind of data based on Discrete Dynamic mechanism are provided
Classification method.
A kind of data classification method based on Discrete Dynamic mechanism, this method comprises:
Set of data samples is received, data prediction is carried out;
Construct dynamic neural network model, every layer of the dynamic neural network model dynamic established including several neurons
Module, and the dynamic module in every layer is connect with a certain dynamic module of next layer, the feature that dynamic module extracts in the last layer
It is fully-connected network between data category;
Using the set of data samples training dynamic neural network model after standardization, data classification model is obtained;
Data to be sorted are received, data classification is carried out according to data classification model.
Further, in the method, the specific steps of the building dynamic neural network model include:
Neuron is modeled as dynamic module, and generates several dynamic modules at random at every layer of deep learning model;
Determine dynamic module next layer of dynamic module connected to it in every layer;
Fully-connected network will be established between the classification of feature and data classification that dynamic module in the last layer extracts, constructed
Dynamic neural network model.
Further, in the method, the specific steps that neuron is modeled as dynamic module include:
Set of data samples is standardized;
Set of data samples after standardization is passed through into dynamic model;
Successively carry out Nonlinear Mapping and pondization processing.
Further, in the method, it is this layer of dynamic that the dynamic model, which is the current state of a certain layer dynamic model,
The current input of model adds the product of this layer of dynamic model parameter current and its eve state.
Further, a certain layer dynamic model parameters are random generation, and meet its characteristic root and be located in unit circle.
Further, in the method, dynamic module is dynamic according to determine the probability next layer connected to it in every layer described
Morphotype block, and the product of the probability and every layer of dynamic module quantity generated at random is greater than 1.
Further, in the method, the fully-connected network is softmax network.
Further, this method further includes the structural parameters using back-propagation algorithm Optimized model, and what is optimized is dynamic
State neural network model.
According to the one aspect of one or more other embodiments of the present disclosure, a kind of computer readable storage medium is provided.
A kind of computer readable storage medium, wherein being stored with a plurality of instruction, described instruction is suitable for by terminal device
Reason device loads and executes a kind of data classification method based on Discrete Dynamic mechanism.
According to the one aspect of one or more other embodiments of the present disclosure, a kind of terminal device is provided.
A kind of terminal device comprising processor and computer readable storage medium, processor is for realizing each instruction;Meter
Calculation machine readable storage medium storing program for executing is suitable for being loaded by processor and being executed described one kind and is based on for storing a plurality of instruction, described instruction
The data classification method of Discrete Dynamic mechanism.
According to the one aspect of one or more other embodiments of the present disclosure, a kind of data based on Discrete Dynamic mechanism are provided
Sorter.
A kind of device for classifying data based on Discrete Dynamic mechanism, based on a kind of number based on Discrete Dynamic mechanism
According to classification method, comprising: sequentially connected data processing module, master cast building module, model training module and data classification
Module;
Data processing module, sample set, carries out data prediction for receiving data;
Model construction module, for constructing dynamic neural network model, if every layer of the dynamic neural network model includes
The dynamic module that dry neuron is established, and the dynamic module in every layer is connect with a certain dynamic module of next layer, in the last layer
It is fully-connected network between the feature that dynamic module extracts and data category;
Model training module, for obtaining using the set of data samples training dynamic neural network model after standardization
To data classification model;
Data categorization module carries out data classification according to data classification model for receiving data to be sorted.
The disclosure the utility model has the advantages that
A kind of data classification method and device based on Discrete Dynamic mechanism that the disclosure provides, uses multilayer dynamic structure
Model, puts forth effort to solve that correlation is underutilized between data in traditional convolution network model and model parameter is excessively asked
Topic, establishes dynamic neural network model, and then training data disaggregated model, the dynamic neural network mould during data classification
Type makes the correlation between data be further enhanced, and model parameter is greatly reduced, and efficiency of algorithm is promoted.The number
Good effect is played not only for the classification of the data such as image, audio according to disaggregated model, is also had very for data such as texts
Strong practical value.
Detailed description of the invention
The accompanying drawings constituting a part of this application is used to provide further understanding of the present application, and the application's shows
Meaning property embodiment and its explanation are not constituted an undue limitation on the present application for explaining the application.
Fig. 1 is a kind of data classification method flow chart based on Discrete Dynamic mechanism according to one or more embodiments;
Fig. 2 is the single dynamic module structural schematic diagram according to one or more embodiments;
Fig. 3 is the dynamic neural network model overall structure diagram according to one or more embodiments;
Fig. 4 is the full connection tier model architecture schematic diagram according to one or more embodiments.
Specific embodiment:
Below in conjunction with the attached drawing in one or more other embodiments of the present disclosure, to one or more other embodiments of the present disclosure
In technical solution be clearly and completely described, it is clear that described embodiments are only a part of the embodiments of the present invention,
Instead of all the embodiments.Based on one or more other embodiments of the present disclosure, those of ordinary skill in the art are not being made
Every other embodiment obtained, shall fall within the protection scope of the present invention under the premise of creative work.
It is noted that following detailed description is all illustrative, it is intended to provide further instruction to the application.Unless another
It indicates, all technical and scientific terms that the present embodiment uses have and the application person of an ordinary skill in the technical field
Normally understood identical meanings.
It should be noted that term used herein above is merely to describe specific embodiment, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singular
Also it is intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet
Include " when, indicate existing characteristics, step, operation, device, component and/or their combination.
It should be noted that flowcharts and block diagrams in the drawings show according to various embodiments of the present disclosure method and
The architecture, function and operation in the cards of system.It should be noted that each box in flowchart or block diagram can represent
A part of one module, program segment or code, a part of the module, program segment or code may include one or more
A executable instruction for realizing the logic function of defined in each embodiment.It should also be noted that some alternately
Realization in, function marked in the box can also occur according to the sequence that is marked in attached drawing is different from.For example, two connect
The box even indicated can actually be basically executed in parallel or they can also be executed in a reverse order sometimes,
This depends on related function.It should also be noted that each box and flow chart in flowchart and or block diagram
And/or the combination of the box in block diagram, the dedicated hardware based system that functions or operations as defined in executing can be used are come
It realizes, or the combination of specialized hardware and computer instruction can be used to realize.
In the absence of conflict, the feature in the embodiment and embodiment in the disclosure can be combined with each other, and tie below
It closes attached drawing and embodiment is described further the disclosure.
According to the one aspect of one or more other embodiments of the present disclosure, a kind of data based on Discrete Dynamic mechanism are provided
Classification method is particularly used in the classification of the data such as text data, image data and audio.
As shown in Figure 1, a kind of data classification method based on Discrete Dynamic mechanism, this method comprises:
S101: receiving set of data samples, carries out data prediction;
S102: building dynamic neural network model, every layer of the dynamic neural network model includes that several neurons are established
Dynamic module, and the dynamic module in every layer is connect with a certain dynamic module of next layer, and dynamic module extracts in the last layer
Feature and data category between be fully-connected network;
S103: using the set of data samples training dynamic neural network model after standardization, data classification mould is obtained
Type;
S104: receiving data to be sorted, carries out data classification according to data classification model.
In the step S102 of the present embodiment, according to the difference of the data such as text data, image data and audio, phase is carried out
The pretreatment such as the data prediction answered, such as filtering, feature extraction, obtains list entries u1,u2,...un。
In the step S102 of the present embodiment, the specific steps of the building dynamic neural network model include:
S1021: neuron is modeled as dynamic module, and generates several dynamic modules at random at every layer of deep learning model
(being set as N number of);
S1022: dynamic module next layer of dynamic module connected to it in every layer is determined;Each module is determined with probability
Whether it is connected with next layer of module;
S1023: fully connected network will be established between the classification of feature and data classification that dynamic module in the last layer extracts
Network constructs dynamic neural network model.
Optimally, the specific steps of the building dynamic neural network model include:
S1021: neuron is modeled as dynamic module, and generates several dynamic modules at random at every layer of deep learning model
(being set as N number of);
S1022: dynamic module next layer of dynamic module connected to it in every layer is determined;Each module is determined with probability
Whether it is connected with next layer of module;
S1023: fully connected network will be established between the classification of feature and data classification that dynamic module in the last layer extracts
Network constructs dynamic neural network model;
S1024: using the structural parameters of back-propagation algorithm Optimized model, final dynamic neural network model is obtained.
In the method, the specific steps that neuron is modeled as dynamic module include:
Set of data samples is standardized;
Set of data samples after standardization is passed through into dynamic model;
Successively carry out Nonlinear Mapping and pondization processing.
Input data is standardized first, standardisation process are as follows:
To list entries u1,u2,...unIt is standardized transformation, the input data after standardizationWherein
The average value of list entriesStandard deviation
It is illustrated in figure 2 individual module structure chart in dynamic neural network model;Nonlinear Mapping is closely followed after dynamic structure
With pond layer, pond block size is 2 × 2.
Further, in the method, it is this layer of dynamic that the dynamic model, which is the current state of a certain layer dynamic model,
The current input of model adds the product of this layer of dynamic model parameter current and its eve state.
Dynamic model are as follows:
x(l)(k+1)=A(l)x(l)(k)+u(l)(k+1),
Wherein A(l)For l layer model parameter, u(l),x(l)Respectively indicate the input and state of l layer model.x(l)(k+1)
For the current state of l layer model, x(l)It (k) is the eve state of l layer model.
The a certain layer dynamic model parameters are random generation, and meet its characteristic root and be located in unit circle.Matrix A(l)For
It is random to generate, and meet characteristic root λ (A(l)) be located in unit circle.
As shown in figure 3, being dynamic neural network model overall architecture.
In the step S102 of the present embodiment, it is every layer described in dynamic module according to determine the probability next layer connected to it
Dynamic module, and the product of the probability and every layer of dynamic module quantity generated at random is greater than 1;Probability p meets pN > 1;Such as Fig. 3
Shown, upper one layer of dynamic module determines whether to be connected with next layer of dynamic module with Probability p.
In the step S102 of the present embodiment, the fully-connected network is softmax network.It is illustrated in figure 4 full connection
Tier model architecture, using between the classification of feature and classification that the last layer is extracted is softmax classifier.
According to the one aspect of one or more other embodiments of the present disclosure, a kind of computer readable storage medium is provided.
A kind of computer readable storage medium, wherein being stored with a plurality of instruction, described instruction is suitable for by terminal device
Reason device loads and executes a kind of data classification method based on Discrete Dynamic mechanism.
According to the one aspect of one or more other embodiments of the present disclosure, a kind of terminal device is provided.
A kind of terminal device comprising processor and computer readable storage medium, processor is for realizing each instruction;Meter
Calculation machine readable storage medium storing program for executing is suitable for being loaded by processor and being executed described one kind and is based on for storing a plurality of instruction, described instruction
The data classification method of Discrete Dynamic mechanism.
These computer executable instructions execute the equipment according to each reality in the disclosure
Apply method or process described in example.
In the present embodiment, computer program product may include computer readable storage medium, containing for holding
The computer-readable program instructions of row various aspects of the disclosure.Computer readable storage medium, which can be, can keep and store
By the tangible device for the instruction that instruction execution equipment uses.Computer readable storage medium for example can be-- but it is unlimited
In-- storage device electric, magnetic storage apparatus, light storage device, electric magnetic storage apparatus, semiconductor memory apparatus or above-mentioned
Any appropriate combination.The more specific example (non exhaustive list) of computer readable storage medium includes: portable computing
Machine disk, hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM or
Flash memory), static random access memory (SRAM), Portable compressed disk read-only memory (CD-ROM), digital versatile disc
(DVD), memory stick, floppy disk, mechanical coding equipment, the punch card for being for example stored thereon with instruction or groove internal projection structure, with
And above-mentioned any appropriate combination.Computer readable storage medium used herein above is not interpreted instantaneous signal itself,
The electromagnetic wave of such as radio wave or other Free propagations, the electromagnetic wave propagated by waveguide or other transmission mediums (for example,
Pass through the light pulse of fiber optic cables) or pass through electric wire transmit electric signal.
Computer-readable program instructions described herein can be downloaded to from computer readable storage medium it is each calculate/
Processing equipment, or outer computer or outer is downloaded to by network, such as internet, local area network, wide area network and/or wireless network
Portion stores equipment.Network may include copper transmission cable, optical fiber transmission, wireless transmission, router, firewall, interchanger, gateway
Computer and/or Edge Server.Adapter or network interface in each calculating/processing equipment are received from network to be counted
Calculation machine readable program instructions, and the computer-readable program instructions are forwarded, for the meter being stored in each calculating/processing equipment
In calculation machine readable storage medium storing program for executing.
Computer program instructions for executing present disclosure operation can be assembly instruction, instruction set architecture (ISA)
Instruction, machine instruction, machine-dependent instructions, microcode, firmware instructions, condition setup data or with one or more programmings
The source code or object code that any combination of language is write, the programming language include the programming language-of object-oriented such as
C++ etc., and conventional procedural programming languages-such as " C " language or similar programming language.Computer-readable program refers to
Order can be executed fully on the user computer, partly be executed on the user computer, as an independent software package
Execute, part on the user computer part on the remote computer execute or completely on a remote computer or server
It executes.In situations involving remote computers, remote computer can include local area network by the network-of any kind
(LAN) or wide area network (WAN)-is connected to subscriber computer, or, it may be connected to outer computer (such as utilize internet
Service provider is connected by internet).In some embodiments, by being believed using the state of computer-readable program instructions
Breath comes personalized customization electronic circuit, such as programmable logic circuit, field programmable gate array (FPGA) or programmable logic
Array (PLA), the electronic circuit can execute computer-readable program instructions, to realize the various aspects of present disclosure.
According to the one aspect of one or more other embodiments of the present disclosure, a kind of data based on Discrete Dynamic mechanism are provided
Sorter.
A kind of device for classifying data based on Discrete Dynamic mechanism, based on a kind of number based on Discrete Dynamic mechanism
According to classification method, comprising: sequentially connected data processing module, master cast building module, model training module and data classification
Module;
Data processing module, sample set, carries out data prediction for receiving data;
Model construction module, for constructing dynamic neural network model, if every layer of the dynamic neural network model includes
The dynamic module that dry neuron is established, and the dynamic module in every layer is connect with a certain dynamic module of next layer, in the last layer
It is fully-connected network between the feature that dynamic module extracts and data category;
Model training module, for obtaining using the set of data samples training dynamic neural network model after standardization
To data classification model;
Data categorization module carries out data classification according to data classification model for receiving data to be sorted.
It should be noted that although being referred to several modules or submodule of equipment in the detailed description above, it is this
Division is only exemplary rather than enforceable.In fact, in accordance with an embodiment of the present disclosure, two or more above-described moulds
The feature and function of block can embody in a module.Conversely, the feature and function of an above-described module can be with
Further division is to be embodied by multiple modules.
The disclosure the utility model has the advantages that
A kind of data classification method and device based on Discrete Dynamic mechanism that the disclosure provides, uses multilayer dynamic structure
Model, puts forth effort to solve that correlation is underutilized between data in traditional convolution network model and model parameter is excessively asked
Topic, establishes dynamic neural network model, and then training data disaggregated model, the dynamic neural network mould during data classification
Type makes the correlation between data be further enhanced, and model parameter is greatly reduced, and efficiency of algorithm is promoted.The number
Good effect is played not only for the classification of the data such as image, audio according to disaggregated model, is also had very for data such as texts
Strong practical value.
The foregoing is merely preferred embodiment of the present application, are not intended to limit this application, for the skill of this field
For art personnel, various changes and changes are possible in this application.Within the spirit and principles of this application, made any to repair
Change, equivalent replacement, improvement etc., should be included within the scope of protection of this application.Therefore, the present invention is not intended to be limited to this
These embodiments shown in text, and it is to fit to the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. a kind of data classification method based on Discrete Dynamic mechanism, which is characterized in that this method comprises:
Set of data samples is received, data prediction is carried out;
Construct dynamic neural network model, every layer of the dynamic neural network model dynamic analog established including several neurons
Block, and the dynamic module in every layer is connect with a certain dynamic module of next layer, the feature that dynamic module extracts in the last layer with
It is fully-connected network between data category;
Using the set of data samples training dynamic neural network model after standardization, data classification model is obtained;
Data to be sorted are received, data classification is carried out according to data classification model.
2. a kind of data classification method based on Discrete Dynamic mechanism as described in claim 1, which is characterized in that in this method
In, the specific steps of the building dynamic neural network model include:
Neuron is modeled as dynamic module, and generates several dynamic modules at random at every layer of deep learning model;
Determine dynamic module next layer of dynamic module connected to it in every layer;
Fully-connected network, building dynamic will be established between the classification of feature and data classification that dynamic module in the last layer extracts
Neural network model.
3. a kind of data classification method based on Discrete Dynamic mechanism as claimed in claim 2, which is characterized in that in this method
In, the specific steps that neuron is modeled as dynamic module include:
Set of data samples is standardized;
Set of data samples after standardization is passed through into dynamic model;
Successively carry out Nonlinear Mapping and pondization processing.
4. a kind of data classification method based on Discrete Dynamic mechanism as claimed in claim 3, which is characterized in that in this method
In, the dynamic model is that the current state of a certain layer dynamic model is the current input of this layer of dynamic model plus this layer of dynamic
The product of model parameter current and its eve state.
Further, a certain layer dynamic model parameters are random generation, and meet its characteristic root and be located in unit circle.
5. a kind of data classification method based on Discrete Dynamic mechanism as described in claim 1, which is characterized in that in this method
In, it is every layer described in dynamic module according to determine the probability next layer of dynamic module connected to it, and the probability and every layer it is random
The product of the dynamic module quantity of generation is greater than 1.
6. a kind of data classification method based on Discrete Dynamic mechanism as described in claim 1, which is characterized in that in this method
In, the fully-connected network is softmax network.
7. a kind of data classification method based on Discrete Dynamic mechanism as described in claim 1, which is characterized in that this method is also
Including the use of the structural parameters of back-propagation algorithm Optimized model, the dynamic neural network model that is optimized.
8. a kind of computer readable storage medium, wherein being stored with a plurality of instruction, which is characterized in that described instruction is suitable for by terminal
The processor of equipment is loaded and is executed such as a kind of described in any item data classifications based on Discrete Dynamic mechanism of claim 1-7
Method.
9. a kind of terminal device comprising processor and computer readable storage medium, processor is for realizing each instruction;It calculates
Machine readable storage medium storing program for executing is for storing a plurality of instruction, which is characterized in that described instruction is suitable for being loaded by processor and being executed such as power
Benefit requires a kind of described in any item data classification methods based on Discrete Dynamic mechanism of 1-7.
10. a kind of device for classifying data based on Discrete Dynamic mechanism, based on such as a kind of described in any item bases of claim 1-7
In the data classification method of Discrete Dynamic mechanism, comprising: sequentially connected data processing module, master cast construct module, model
Training module and data categorization module;
Data processing module, sample set, carries out data prediction for receiving data;
Model construction module, for constructing dynamic neural network model, every layer of the dynamic neural network model includes several minds
The dynamic module established through member, and the dynamic module in every layer is connect with a certain dynamic module of next layer, dynamic in the last layer
It is fully-connected network between the feature that module is extracted and data category;
Model training module, for being counted using the set of data samples training dynamic neural network model after standardization
According to disaggregated model;
Data categorization module carries out data classification according to data classification model for receiving data to be sorted.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910122983.9A CN109919203A (en) | 2019-02-19 | 2019-02-19 | A kind of data classification method and device based on Discrete Dynamic mechanism |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910122983.9A CN109919203A (en) | 2019-02-19 | 2019-02-19 | A kind of data classification method and device based on Discrete Dynamic mechanism |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109919203A true CN109919203A (en) | 2019-06-21 |
Family
ID=66961743
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910122983.9A Pending CN109919203A (en) | 2019-02-19 | 2019-02-19 | A kind of data classification method and device based on Discrete Dynamic mechanism |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109919203A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110807497A (en) * | 2019-10-10 | 2020-02-18 | 山东师范大学 | Handwritten data classification method and system based on deep dynamic network |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107169566A (en) * | 2017-06-09 | 2017-09-15 | 山东师范大学 | Dynamic neural network model training method and device |
CN107480774A (en) * | 2017-08-11 | 2017-12-15 | 山东师范大学 | Dynamic neural network model training method and device based on integrated study |
-
2019
- 2019-02-19 CN CN201910122983.9A patent/CN109919203A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107169566A (en) * | 2017-06-09 | 2017-09-15 | 山东师范大学 | Dynamic neural network model training method and device |
CN107480774A (en) * | 2017-08-11 | 2017-12-15 | 山东师范大学 | Dynamic neural network model training method and device based on integrated study |
Non-Patent Citations (2)
Title |
---|
GYÖRGY KOVÁCS ET AL.: "Increasing the robustness of CNN acoustic models using autoregressive moving average spectrogram features and channel dropout", 《PATTERN RECOGNITION LETTERS》 * |
任晓霞: "基于Dropout深度卷积神经网络的ST段波形分类算法", 《传感技术学报》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110807497A (en) * | 2019-10-10 | 2020-02-18 | 山东师范大学 | Handwritten data classification method and system based on deep dynamic network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11783227B2 (en) | Method, apparatus, device and readable medium for transfer learning in machine learning | |
CN110188919A (en) | A kind of load forecasting method based on shot and long term memory network | |
CN113191479B (en) | Method, system, node and storage medium for joint learning | |
CN112580328A (en) | Event information extraction method and device, storage medium and electronic equipment | |
CN111310918B (en) | Data processing method, device, computer equipment and storage medium | |
Zeb et al. | Edge intelligence in softwarized 6G: Deep learning-enabled network traffic predictions | |
CN110288024A (en) | A kind of Image Classifier building, image-recognizing method and system for lacking sample learning based on prototype network | |
CN110188210A (en) | One kind is based on figure regularization and the independent cross-module state data retrieval method of mode and system | |
CN115114409B (en) | Civil aviation unsafe event combined extraction method based on soft parameter sharing | |
CN109978074A (en) | Image aesthetic feeling and emotion joint classification method and system based on depth multi-task learning | |
CN114282059A (en) | Video retrieval method, device, equipment and storage medium | |
CN110046279A (en) | Prediction technique, medium, device and the calculating equipment of video file feature | |
Ren et al. | A novel deep learning method for application identification in wireless network | |
CN116720132A (en) | Power service identification system, method, device, medium and product | |
KR102002549B1 (en) | Method and apparatus for generating multi-stage classification model | |
CN106649743A (en) | Method and system for storing and sharing creative idea classified brain library resources | |
CN110163255A (en) | A kind of data stream clustering method and device based on density peaks | |
CN109919203A (en) | A kind of data classification method and device based on Discrete Dynamic mechanism | |
CN111445545A (en) | Text-to-map method, device, storage medium and electronic equipment | |
CN114900435B (en) | Connection relation prediction method and related equipment | |
CN116911757A (en) | Service realization method, device and storage medium | |
CN109978013A (en) | A kind of depth clustering method for figure action identification | |
CN113747480B (en) | Processing method and device for 5G slice faults and computing equipment | |
CN115587616A (en) | Network model training method and device, storage medium and computer equipment | |
CN114756425A (en) | Intelligent monitoring method and device, electronic equipment and computer readable storage medium |
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: 20190621 |