CN114492789B - Neural network model construction method and device for data samples - Google Patents
Neural network model construction method and device for data samples Download PDFInfo
- Publication number
- CN114492789B CN114492789B CN202210085874.6A CN202210085874A CN114492789B CN 114492789 B CN114492789 B CN 114492789B CN 202210085874 A CN202210085874 A CN 202210085874A CN 114492789 B CN114492789 B CN 114492789B
- Authority
- CN
- China
- Prior art keywords
- neural network
- layer
- network
- network model
- input
- 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.)
- Active
Links
- 238000003062 neural network model Methods 0.000 title claims abstract description 120
- 238000010276 construction Methods 0.000 title claims abstract description 10
- 210000002569 neuron Anatomy 0.000 claims abstract description 85
- 238000000034 method Methods 0.000 claims abstract description 47
- 238000012549 training Methods 0.000 claims abstract description 41
- 238000013507 mapping Methods 0.000 claims abstract description 10
- 230000009467 reduction Effects 0.000 claims abstract description 9
- 239000010410 layer Substances 0.000 claims description 212
- 238000013528 artificial neural network Methods 0.000 claims description 100
- 230000006870 function Effects 0.000 claims description 22
- 238000012545 processing Methods 0.000 claims description 22
- 230000004913 activation Effects 0.000 claims description 16
- 238000010606 normalization Methods 0.000 claims description 10
- 239000002356 single layer Substances 0.000 claims description 9
- 230000015654 memory Effects 0.000 description 17
- 230000008569 process Effects 0.000 description 10
- 238000012360 testing method Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 230000005291 magnetic effect Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000003909 pattern recognition Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- KLDZYURQCUYZBL-UHFFFAOYSA-N 2-[3-[(2-hydroxyphenyl)methylideneamino]propyliminomethyl]phenol Chemical compound OC1=CC=CC=C1C=NCCCN=CC1=CC=CC=C1O KLDZYURQCUYZBL-UHFFFAOYSA-N 0.000 description 1
- 230000003213 activating effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 201000001098 delayed sleep phase syndrome Diseases 0.000 description 1
- 208000033921 delayed sleep phase type circadian rhythm sleep disease Diseases 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000005294 ferromagnetic effect Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G06N3/08—Learning methods
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Neurology (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a neural network model construction method and device for a data sample, comprising the following steps: the method comprises the steps that a preset neural network model is obtained, the preset neural network model comprises an input layer, a projection network and a learning network, the input layer is used for obtaining input sample signals and mapping the input sample signals into neuron outputs, the projection network is used for reducing dimensions of high-dimensional information output by the input layer, and the learning network is used for learning the information subjected to dimension reduction by the projection network and outputting a learning result; determining a connection weight of each layer of the preset neural network model based on a target task corresponding to the input sample signal; training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model. According to the invention, the projection network is constructed, so that the projection study on the data sample can be facilitated, the information can be rapidly transmitted in the network, and the efficient study on the sample can be realized.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a method and an apparatus for constructing a neural network model of a data sample.
Background
The artificial neural network is an important research direction in the field of artificial intelligence, and is a simulation of the human brain neural network from the point of information processing. The artificial neural network can be arbitrarily approximated to any nonlinear function, so that the distinction of different modes of signals is realized. Artificial neural networks have been widely used in many fields such as signal processing, pattern recognition, automatic control, and artificial intelligence. Many models exist for artificial neural networks, such as BP model, hopfield model, overrun learning machine, deep learning model, etc. These models are typically composed of an input layer, a hidden layer, and an output layer. Information is handled and transferred between layers. In each layer, the neurons perform nonlinear processing on the input signals through an activation function, so that feature extraction is realized. Each layer is made up of a plurality of neuronal nodes. Adjacent layer neuron nodes are connected with each other, but common layer neuron nodes and cross-layer neuron nodes are not connected with each other. The connection between the neurons has weight, and the learning process is a process of continuously modifying the connection weight between the neurons. The BP neural network adjusts the weight through forward transmission of information and reverse transmission of errors.
Based on BP neural network, deep learning re-activates the research of neural network, which pushes artificial neural network from shallow layer to deep layer, and opens new era of Deep Neural Network (DNN). The deep neural network has more hidden layers, the first hidden layer extracts basic features from the original data, and the later hidden layers combine the basic features into one higher-order abstract feature. Deep learning can automatically extract the characteristics required by classification, does not need human participation, and has achieved great success in the fields of voice recognition, image processing, pattern recognition and the like.
Deep learning, however, requires a significant amount of time to complete the training of the network, as with previous neural networks. Although in the big data age, expert knowledge samples are scarce. On the other hand, the computing power in practical application is often limited, the performance of DNN is restricted by limited computing resources, and meanwhile, the energy and benefits can be saved due to the saving of computing resources. Therefore, an artificial neural network having a strong generalization ability and a high learning speed needs to be studied.
Disclosure of Invention
Aiming at the problems, the invention provides a method and a device for constructing a neural network model of a data sample, which realize the rapid transmission of information in a network and improve the learning efficiency of sample data.
In order to achieve the above object, the present invention provides the following technical solutions:
A neural network model construction method of a data sample comprises the following steps:
The method comprises the steps that a preset neural network model is obtained, the preset neural network model comprises an input layer, a projection network and a learning network, the input layer is used for obtaining input sample signals and mapping the input sample signals into neuron outputs, the projection network is used for reducing dimensions of high-dimensional information output by the input layer, and the learning network is used for learning the information subjected to dimension reduction by the projection network and outputting a learning result;
Determining a connection weight of each layer of the preset neural network model based on a target task corresponding to the input sample signal;
Training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model.
Optionally, the method further comprises:
obtaining a target training sample;
and determining the quantity of neurons of each layer of the preset neural network model based on the length of the target training sample.
Optionally, the input layer of the preset neural network model is composed of a single-layer neural network; the projection network consists of a plurality of layers of neural networks, the layers of neural networks of the projection network have specific connection weights, each layer of neural network consists of a plurality of neurons, and each layer of neural network is connected with only a limited number of neurons of the previous layer; the learning network is composed of a shallow neural network, a first layer neural network of the neural network is connected with a first layer neural network of the projection network, the shallow neural network is a full-connection network, and neurons of different layers of the shallow neural network are connected through weights.
Optionally, the determining, based on the target task corresponding to the input sample signal, a connection weight of each layer of the preset neural network model includes:
Acquiring a target training sample corresponding to the input sample signal;
Inputting the target training sample to the input layer, and obtaining the output of neurons of the input layer;
Inputting the output of the input layer neuron into the projection network to obtain an output signal of the projection network;
inputting the output signal of the projection network to the learning network, and obtaining the learning result of the learning network;
and adjusting the connection weight of the neuron based on the comparison value of the target task and the learning result to obtain the connection weight of each layer of the preset neural network model.
Optionally, the method further comprises:
determining an activation function of each layer of the neural network model when determining output information of each layer of the preset neural network model, and carrying out normalization processing on signals input by each layer;
And determining the output signal of each layer according to the input signals after normalization processing and the corresponding activation functions.
A neural network model building apparatus for data samples, comprising:
The acquisition unit is used for acquiring a preset neural network model, the preset neural network model comprises an input layer, a projection network and a learning network, the input layer is used for acquiring input sample signals and mapping the input sample signals into neuron outputs, the projection network is used for reducing the dimension of high-dimension information output by the input layer, and the learning network is used for learning the dimension reduced information of the projection network and outputting a learning result;
The determining unit is used for determining a connection weight value of each layer of the preset neural network model based on a target task corresponding to the input sample signal;
the training unit is used for training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model.
Optionally, the method further comprises:
the neuron number determining unit is used for obtaining a target training sample;
and determining the quantity of neurons of each layer of the preset neural network model based on the length of the target training sample.
Optionally, the input layer of the preset neural network model is composed of a single-layer neural network; the projection network consists of a plurality of layers of neural networks, the layers of neural networks of the projection network have specific connection weights, each layer of neural network consists of a plurality of neurons, and each layer of neural network is connected with only a limited number of neurons of the previous layer; the learning network is composed of a shallow neural network, a first layer neural network of the neural network is connected with a first layer neural network of the projection network, the shallow neural network is a full-connection network, and neurons of different layers of the shallow neural network are connected through weights.
Optionally, the determining unit is specifically configured to:
Acquiring a target training sample corresponding to the input sample signal;
Inputting the target training sample to the input layer, and obtaining the output of neurons of the input layer;
Inputting the output of the input layer neuron into the projection network to obtain an output signal of the projection network;
inputting the output signal of the projection network to the learning network, and obtaining the learning result of the learning network;
and adjusting the connection weight of the neuron based on the comparison value of the target task and the learning result to obtain the connection weight of each layer of the preset neural network model.
Optionally, the apparatus further comprises:
An output signal determining unit, configured to determine an activation function of each layer of the neural network model and normalize signals input by each layer when determining output information of each layer of the preset neural network model;
And determining the output signal of each layer according to the input signals after normalization processing and the corresponding activation functions.
Compared with the prior art, the invention provides a neural network model construction method and device for a data sample, comprising the following steps: the method comprises the steps that a preset neural network model is obtained, the preset neural network model comprises an input layer, a projection network and a learning network, the input layer is used for obtaining input sample signals and mapping the input sample signals into neuron outputs, the projection network is used for reducing dimensions of high-dimensional information output by the input layer, and the learning network is used for learning the information subjected to dimension reduction by the projection network and outputting a learning result; determining a connection weight of each layer of the preset neural network model based on a target task corresponding to the input sample signal; training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model. According to the invention, the projection network is constructed, so that the projection study on the data sample can be facilitated, the information can be rapidly transmitted in the network, and the efficient study on the sample can be realized.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic flow chart of a neural network model construction method of a data sample according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a projection learning model according to an embodiment of the present invention;
Fig. 3 is a schematic structural diagram of a neural network model building device for data samples according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The terms first and second and the like in the description and in the claims and in the above-described figures are used for distinguishing between different objects and not necessarily for describing a sequential or chronological order. Furthermore, the terms "comprise" and "have," as well as any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to the listed steps or elements but may include steps or elements not expressly listed.
In an embodiment of the present invention, a method for constructing a neural network model of a data sample is provided, referring to fig. 1, the method may include the following steps:
S101, acquiring a preset neural network model.
The preset neural network model comprises an input layer, a projection network and a learning network, wherein the input layer is used for acquiring input sample signals and mapping the input sample signals into nerve cells for output, the projection network is used for reducing the dimension of high-dimension information output by the input layer, and the learning network is used for learning the dimension reduced information of the projection network and outputting a learning result.
Specifically, the input layer of the preset neural network model consists of a single-layer neural network; the projection network consists of a plurality of layers of neural networks, the layers of neural networks of the projection network have specific connection weights, each layer of neural network consists of a plurality of neurons, and each layer of neural network is connected with only a limited number of neurons of the previous layer; the learning network is composed of a shallow neural network, a first layer neural network of the neural network is connected with a first layer neural network of the projection network, the shallow neural network is a full-connection network, and neurons of different layers of the shallow neural network are connected through weights.
S102, determining a connection weight of each layer of the preset neural network model based on a target task corresponding to the input sample signal.
The target task may be any target task that is desired to be achieved in the training process based on the target training sample corresponding to the input sample signal, which is not limited.
The final output signal can be obtained by processing each layer in the preset neural network model based on the output signal of the previous layer, then the output signal is compared with the expected output signal of the target task, and the connection weight of each layer is adjusted based on the comparison result so as to obtain the connection weight of each layer corresponding to the target task.
In one implementation manner of the embodiment of the present application, the determining, based on the target task corresponding to the input sample signal, a connection weight of each layer of the preset neural network model includes:
Acquiring a target training sample corresponding to the input sample signal;
Inputting the target training sample to the input layer, and obtaining the output of neurons of the input layer;
Inputting the output of the input layer neuron into the projection network to obtain an output signal of the projection network;
inputting the output signal of the projection network to the learning network, and obtaining the learning result of the learning network;
and adjusting the connection weight of the neuron based on the comparison value of the target task and the learning result to obtain the connection weight of each layer of the preset neural network model.
Further, the method further comprises:
determining an activation function of each layer of the neural network model when determining output information of each layer of the preset neural network model, and carrying out normalization processing on signals input by each layer;
And determining the output signal of each layer according to the input signals after normalization processing and the corresponding activation functions.
And S103, training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model.
After the connection weight of each layer is obtained, the association weight between the preset neural network models is adjusted and trained, and the target neural network model is obtained. The target neural network can be used for processing data in a subsequent application scenario similar to the target task.
The embodiment of the application provides a neural network model construction method of a data sample, which comprises the following steps: the method comprises the steps that a preset neural network model is obtained, the preset neural network model comprises an input layer, a projection network and a learning network, the input layer is used for obtaining input sample signals and mapping the input sample signals into neuron outputs, the projection network is used for reducing dimensions of high-dimensional information output by the input layer, and the learning network is used for learning the information subjected to dimension reduction by the projection network and outputting a learning result; determining a connection weight of each layer of the preset neural network model based on a target task corresponding to the input sample signal; training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model. According to the application, the projection network is constructed, so that the projection study on the data sample can be facilitated, the information can be rapidly transmitted in the network, and the efficient study on the sample can be realized.
The preset neural network model provided in the embodiment of the application is a projection learning model, and comprises an input layer, a projection network and a learning network.
The input layer consists of a single-layer neural network and is used for receiving an input sample signal and mapping the input sample signal into the output of neurons; the single-layer neural network consists of a plurality of neurons, and the neurons of the single-layer neural network are not connected.
The projection network consists of a plurality of layers of neural networks and is used for rapidly projecting the output of the input layer to the learning network so as to realize the dimension reduction of the high-dimensional information of the input layer; the projection network comprises a plurality of layers of neural networks, wherein the neural networks are provided with specific connection weights, the weights do not need to be modified in a learning process, and each layer of neural network consists of a plurality of neurons; the projection network neurons have a local field of view, that is, the projection network neurons are connected with only a limited number of neurons in the previous layer; the local view field can be dynamically adjusted.
The learning network consists of a shallow neural network and is used for learning and memorizing the information of the projected network after dimension reduction and outputting the learning and memorizing result; the first layer of the neural network of the learning network is connected with the last layer of the neural network of the projection network; the shallow neural network is that the number of layers of the neural network is generally not more than five; the shallow neural network is a fully connected network; neurons among different layers of the shallow neural network are connected through weights, and the learning process is a process of modifying the connection weights of the shallow neural network; and the shallow neural network outputs a learning result in the last layer.
Embodiments of the present application will be described below with a process of learning one-dimensional data samples. For example, the CWRU data is taken as an example, the CWRU data is measured by the bearing data center of kesixi Chu Da, usa, and the processing may be performed based on other data.
Firstly, CWRU bearing data in a time domain are transformed into the frequency domain, and a training sample library is established. Constructing a projection learning network model according to the length of the sample; the sample length selected here is 2048, and the constructed projection learning model is shown in fig. 2, where n=2048 and p=d=16. In this way, the designed projection learning model has an input layer composed of 2048 neurons, a projection network composed of two layers of neural networks, a first projection layer composed of 2048 neurons, a second projection layer composed of 128 neurons, and a learning network composed of three layers of neural networks, wherein the number of neurons of the first learning layer is the same as that of neurons of the second projection layer, namely 128 neurons of the first learning layer, 32 neurons of the second learning layer, and the number of neurons of the third learning layer is determined by the category number of the sample to be learned.
The input sample data is mapped to the output of neurons through the input layers and serves as the input signal to the projection network, and the output of each input layer neuron can be expressed as:
yi＝f(xi) (1)
Wherein x i is a normalized sample signal, f (·) represents an activation function, the input layer employs a Tanh function whose activation function is optimized, and the expression is Each neuron of the first projection network receives output signals of 16 input layer neurons, the outputs of which are:
In the middle of Called coefficient of action, selected here/>The activating function is a sigmoid function, and the expression is
Each neuron of the second projection network receives output signals of 16 first projection network neurons, and outputs of the output signals are as follows:
In the middle of Also referred to as the coefficient of action, also referred to herein as/>The activation function is selected as the same as the first projected network.
The learning network consists of three layers of neural networks, wherein the first layer of learning network is connected with the second layer of projection network, and the first layer of learning network neurons receive output signals of the second layer of projection network neurons. The neurons among the three learning layers are connected through weights, and the learning process is the process of modifying the connecting weights of the three layers of neural networks. The result of the learning training is output by the third layer learning network. It can be seen that the projection learning model constructed by the test of the invention consists of a 6-layer network.
The projection learning model was tested using CWRU bearing data and compared to a deep neural network model, which was designed to be 5 layers for ease of comparison, with 2048-414-114-45-N where N represents the number of output layer neurons. The comparison results are shown in Table 1. From this table, the generalization ability of the method of the present invention is much improved over the original deep neural network model. The identification rate of the method in the first test and the third test reaches 100%, while the original method only has 85.31% and 90.45%, and the identification rate of the method is improved by at least 9.55%. The learning time of the two methods is shown in table 2, and the training time of the method is greatly reduced and is less than 2% of the training time of the original neural network model. This shows that the method of the invention has not only very strong generalization ability, but also high learning speed.
Table 1 test comparison results of recognition rate
In table 1, F represents a failure sample, and N represents a normal sample.
Table 2 test comparison results of learning time
The neural network model construction method for the data samples can realize rapid learning of the data samples, save a great deal of time consumed for training the traditional neural network model and save computing resources. Compared with the traditional neural network model, the method has better generalization capability and quicker learning and training speed, and has wide industrial application prospect.
In another embodiment of the present application, there is also provided a neural network model building apparatus for data samples, referring to fig. 3, including:
the acquiring unit 301 is configured to acquire a preset neural network model, where the preset neural network model includes an input layer, a projection network, and a learning network, the input layer is configured to acquire an input sample signal and map the input sample signal into a neuron for output, the projection network is configured to perform dimension reduction on high-dimensional information output by the input layer, and the learning network is configured to learn dimension-reduced information of the projection network and output a learning result;
A determining unit 302, configured to determine a connection weight value of each layer of the preset neural network model based on a target task corresponding to the input sample signal;
And the training unit 303 is configured to train the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model.
Optionally, the method further comprises:
the neuron number determining unit is used for obtaining a target training sample;
and determining the quantity of neurons of each layer of the preset neural network model based on the length of the target training sample.
Optionally, the input layer of the preset neural network model is composed of a single-layer neural network; the projection network consists of a plurality of layers of neural networks, the layers of neural networks of the projection network have specific connection weights, each layer of neural network consists of a plurality of neurons, and each layer of neural network is connected with only a limited number of neurons of the previous layer; the learning network is composed of a shallow neural network, a first layer neural network of the neural network is connected with a first layer neural network of the projection network, the shallow neural network is a full-connection network, and neurons of different layers of the shallow neural network are connected through weights.
Optionally, the determining unit is specifically configured to:
Acquiring a target training sample corresponding to the input sample signal;
Inputting the target training sample to the input layer, and obtaining the output of neurons of the input layer;
Inputting the output of the input layer neuron into the projection network to obtain an output signal of the projection network;
inputting the output signal of the projection network to the learning network, and obtaining the learning result of the learning network;
and adjusting the connection weight of the neuron based on the comparison value of the target task and the learning result to obtain the connection weight of each layer of the preset neural network model.
Optionally, the apparatus further comprises:
An output signal determining unit, configured to determine an activation function of each layer of the neural network model and normalize signals input by each layer when determining output information of each layer of the preset neural network model;
And determining the output signal of each layer according to the input signals after normalization processing and the corresponding activation functions.
The embodiment of the application provides a neural network model construction device of a data sample, which comprises the following components: the method comprises the steps that a preset neural network model is obtained, the preset neural network model comprises an input layer, a projection network and a learning network, the input layer is used for obtaining input sample signals and mapping the input sample signals into neuron outputs, the projection network is used for reducing dimensions of high-dimensional information output by the input layer, and the learning network is used for learning the information subjected to dimension reduction by the projection network and outputting a learning result; determining a connection weight of each layer of the preset neural network model based on a target task corresponding to the input sample signal; training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model. According to the application, the projection network is constructed, so that the projection study on the data sample can be facilitated, the information can be rapidly transmitted in the network, and the efficient study on the sample can be realized.
Based on the foregoing embodiments, embodiments of the present application provide a computer-readable storage medium storing one or more programs executable by one or more processors to implement the steps of the neural network model building method of the data samples of any one of the above.
The embodiment of the invention also provides electronic equipment, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the steps of the neural network model building method of the data samples.
The Processor or CPU may be at least one of an Application SPECIFIC INTEGRATED Circuit (ASIC), a digital signal Processor (DIGITAL SIGNAL Processor, DSP), a digital signal processing device (DIGITAL SIGNAL Processing Device, DSPD), a programmable logic device (Programmable Logic Device, PLD), a field programmable gate array (Field Programmable GATE ARRAY, FPGA), a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, and a microprocessor. It will be appreciated that the electronic device implementing the above-mentioned processor function may be other, and embodiments of the present application are not limited in detail.
The computer storage medium/Memory may be a Read Only Memory (ROM), a programmable Read Only Memory (Programmable Read-Only Memory, PROM), an erasable programmable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), an electrically erasable programmable Read Only Memory (ELECTRICALLY ERASABLE PROGRAMMABLE READ-Only Memory, EEPROM), a magnetic random access Memory (Ferromagnetic Random Access Memory, FRAM), a Flash Memory (Flash Memory), a magnetic surface Memory, an optical disk, or a compact disk Read Only Memory (Compact Disc Read-Only Memory, CD-ROM), or the like; but may also be various terminals such as mobile phones, computers, tablet devices, personal digital assistants, etc., that include one or any combination of the above-mentioned memories.
In the several embodiments provided by the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above described device embodiments are only illustrative, e.g. the division of the units is only one logical function division, and there may be other divisions in practice, such as: multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the various components shown or discussed may be coupled or directly coupled or communicatively coupled to each other via some interface, whether indirectly coupled or communicatively coupled to devices or units, whether electrically, mechanically, or otherwise.
The units described as separate units may or may not be physically separate, and units displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing module, or each unit may be separately used as one unit, or two or more units may be integrated in one unit; the integrated units may be implemented in hardware or in hardware plus software functional units. Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware associated with program instructions, where the foregoing program may be stored in a computer readable storage medium, and when executed, the program performs steps including the above method embodiments; and the aforementioned storage medium includes: a mobile storage device, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or a optical disk, or the like, which can store program codes.
The methods disclosed in the method embodiments provided by the application can be arbitrarily combined under the condition of no conflict to obtain a new method embodiment.
The features disclosed in the several product embodiments provided by the application can be combined arbitrarily under the condition of no conflict to obtain new product embodiments.
The features disclosed in the embodiments of the method or the apparatus provided by the application can be arbitrarily combined without conflict to obtain new embodiments of the method or the apparatus.
The foregoing is merely illustrative of the present application, and the present application is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (2)
1. A method for constructing a neural network model of a data sample, comprising:
The method comprises the steps that a preset neural network model is obtained, the preset neural network model comprises an input layer, a projection network and a learning network, the input layer is used for obtaining input sample signals and mapping the input sample signals into neuron outputs, the projection network is used for reducing dimensions of high-dimensional information output by the input layer, and the learning network is used for learning the information subjected to dimension reduction by the projection network and outputting a learning result; the neuron number of each layer of the preset neural network model is determined based on the length of the target training sample; the input layer of the preset neural network model consists of a single-layer neural network; the projection network consists of a plurality of layers of neural networks, the layers of neural networks of the projection network have specific connection weights, each layer of neural network consists of a plurality of neurons, and each layer of neural network is connected with only a limited number of neurons of the previous layer; the learning network consists of a shallow neural network, a first layer neural network of the neural network is connected with a first layer neural network of the projection network, the shallow neural network is a full-connection network, and neurons of different layers of the shallow neural network are connected through weights; further comprises: determining an activation function of each layer of the neural network model when determining output information of each layer of the preset neural network model, and carrying out normalization processing on signals input by each layer; determining an output signal of each layer according to the input signals after normalization processing and the corresponding activation functions;
Determining a connection weight of each layer of the preset neural network model based on a target task corresponding to the input sample signal, wherein the determining the connection weight of each layer of the preset neural network model based on the target task corresponding to the input sample signal includes: acquiring a target training sample corresponding to the input sample signal; inputting the target training sample to the input layer, and obtaining the output of neurons of the input layer; inputting the output of the input layer neuron into the projection network to obtain an output signal of the projection network; inputting the output signal of the projection network to the learning network, and obtaining the learning result of the learning network; based on the comparison value of the target task and the learning result, adjusting the connection weight of the neuron to obtain the connection weight of each layer of the preset neural network model;
Training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model.
2. A neural network model construction device for a data sample, comprising:
The acquisition unit is used for acquiring a preset neural network model, the preset neural network model comprises an input layer, a projection network and a learning network, the input layer is used for acquiring input sample signals and mapping the input sample signals into neuron outputs, the projection network is used for reducing the dimension of high-dimension information output by the input layer, and the learning network is used for learning the dimension reduced information of the projection network and outputting a learning result; the neuron number of each layer of the preset neural network model is determined based on the length of the target training sample; the input layer of the preset neural network model consists of a single-layer neural network; the projection network consists of a plurality of layers of neural networks, the layers of neural networks of the projection network have specific connection weights, each layer of neural network consists of a plurality of neurons, and each layer of neural network is connected with only a limited number of neurons of the previous layer; the learning network consists of a shallow neural network, a first layer neural network of the neural network is connected with a first layer neural network of the projection network, the shallow neural network is a full-connection network, and neurons of different layers of the shallow neural network are connected through weights; further comprises: determining an activation function of each layer of the neural network model when determining output information of each layer of the preset neural network model, and carrying out normalization processing on signals input by each layer; determining an output signal of each layer according to the input signals after normalization processing and the corresponding activation functions;
The determining unit is used for determining a connection weight value of each layer of the preset neural network model based on a target task corresponding to the input sample signal; the determining unit is specifically configured to obtain a target training sample corresponding to the input sample signal; inputting the target training sample to the input layer, and obtaining the output of neurons of the input layer; inputting the output of the input layer neuron into the projection network to obtain an output signal of the projection network; inputting the output signal of the projection network to the learning network, and obtaining the learning result of the learning network; based on the comparison value of the target task and the learning result, adjusting the connection weight of the neuron to obtain the connection weight of each layer of the preset neural network model;
the training unit is used for training the preset neural network model according to the connection weight of each layer of the preset neural network model to obtain a target neural network model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210085874.6A CN114492789B (en) | 2022-01-25 | 2022-01-25 | Neural network model construction method and device for data samples |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210085874.6A CN114492789B (en) | 2022-01-25 | 2022-01-25 | Neural network model construction method and device for data samples |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114492789A CN114492789A (en) | 2022-05-13 |
| CN114492789B true CN114492789B (en) | 2024-05-14 |
Family
ID=81475291
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210085874.6A Active CN114492789B (en) | 2022-01-25 | 2022-01-25 | Neural network model construction method and device for data samples |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114492789B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN106295799A (en) * | 2015-05-12 | 2017-01-04 | 核工业北京地质研究院 | A kind of implementation method of degree of depth study multilayer neural network |
| CN107301453A (en) * | 2016-04-15 | 2017-10-27 | 北京中科寒武纪科技有限公司 | The artificial neural network forward operation apparatus and method for supporting discrete data to represent |
| CN110197260A (en) * | 2019-06-06 | 2019-09-03 | 百度在线网络技术（北京）有限公司 | A kind of data processing method and device |
| CN111626994A (en) * | 2020-05-18 | 2020-09-04 | 江苏远望仪器集团有限公司 | Equipment fault defect diagnosis method based on improved U-Net neural network |
-
2022
- 2022-01-25 CN CN202210085874.6A patent/CN114492789B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106295799A (en) * | 2015-05-12 | 2017-01-04 | 核工业北京地质研究院 | A kind of implementation method of degree of depth study multilayer neural network |
| CN107301453A (en) * | 2016-04-15 | 2017-10-27 | 北京中科寒武纪科技有限公司 | The artificial neural network forward operation apparatus and method for supporting discrete data to represent |
| 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 |
| CN110197260A (en) * | 2019-06-06 | 2019-09-03 | 百度在线网络技术（北京）有限公司 | A kind of data processing method and device |
| CN111626994A (en) * | 2020-05-18 | 2020-09-04 | 江苏远望仪器集团有限公司 | Equipment fault defect diagnosis method based on improved U-Net neural network |
Non-Patent Citations (1)
| Title |
|---|
| " 大功率盘型激光焊熔池形态在线识别及焊缝成形预测模型";张艳喜;《中国博士学位论文全文数据库 信息科技辑》;20150315(第03期);第I138-26页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114492789A (en) | 2022-05-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110799994B (en) | Adaptive bit width reduction for neural networks | |
| US10311326B2 (en) | Systems and methods for improved image textures | |
| EP3937087A1 (en) | Knowledge-graph biased classification for data | |
| CN109522945B (en) | Group emotion recognition method and device, intelligent device and storage medium | |
| CN108960314B (en) | Training method and device based on difficult samples and electronic equipment | |
| CN109886422A (en) | Model configuration method, device, electronic equipment and read/write memory medium | |
| US20200167593A1 (en) | Dynamic reconfiguration training computer architecture | |
| CN107223260B (en) | Method for dynamically updating classifier complexity | |
| US11080592B2 (en) | Neuromorphic architecture for feature learning using a spiking neural network | |
| CN114595874A (en) | Ultra-short-term power load prediction method based on dynamic neural network | |
| CN106778910A (en) | Deep learning system and method based on local training | |
| CN111357051A (en) | Speech emotion recognition method, intelligent device and computer readable storage medium | |
| CN114332545A (en) | Image data classification method and device based on low-bit pulse neural network | |
| CN112749737A (en) | Image classification method and device, electronic equipment and storage medium | |
| Sun et al. | Low-consumption neuromorphic memristor architecture based on convolutional neural networks | |
| US11195053B2 (en) | Computer architecture for artificial image generation | |
| US9536190B2 (en) | Dynamically assigning and examining synaptic delay | |
| CN116126354A (en) | Model deployment method, device, electronic equipment and storage medium | |
| US9542645B2 (en) | Plastic synapse management | |
| Li et al. | Disease identification in potato leaves using swin transformer | |
| CN114492789B (en) | Neural network model construction method and device for data samples | |
| WO2021179198A1 (en) | Image feature visualization method, image feature visualization apparatus, and electronic device | |
| CN111507407A (en) | Training method and device of image classification model | |
| CN112183725B (en) | Method of providing neural network, computing device, and computer-readable storage medium | |
| CN111898465B (en) | Method and device for acquiring face recognition model |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |