CN112932497A - Unbalanced single-lead electrocardiogram data classification method and system - Google Patents
Unbalanced single-lead electrocardiogram data classification method and system Download PDFInfo
- Publication number
- CN112932497A CN112932497A CN202110258782.9A CN202110258782A CN112932497A CN 112932497 A CN112932497 A CN 112932497A CN 202110258782 A CN202110258782 A CN 202110258782A CN 112932497 A CN112932497 A CN 112932497A
- Authority
- CN
- China
- Prior art keywords
- features
- class
- minority
- electrocardiosignal
- original
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000000605 extraction Methods 0.000 claims abstract description 17
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 15
- 238000003786 synthesis reaction Methods 0.000 claims abstract description 15
- 238000007781 pre-processing Methods 0.000 claims abstract description 14
- 230000002194 synthesizing effect Effects 0.000 claims abstract description 10
- 238000013527 convolutional neural network Methods 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 9
- 238000000354 decomposition reaction Methods 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7253—Details of waveform analysis characterised by using transforms
- A61B5/726—Details of waveform analysis characterised by using transforms using Wavelet transforms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a method and a system for classifying unbalanced single-lead electrocardiogram data, wherein the method comprises the following steps: acquiring an original electrocardiosignal and preprocessing the original electrocardiosignal to obtain a preprocessed electrocardiosignal; carrying out depth feature extraction on the preprocessed electrocardiosignals and reducing dimensions of the depth features to obtain a majority class of features and a minority class of features; synthesizing the minority class of features based on the SMOTE algorithm to obtain a minority class of feature synthesis sample; and performing SVM classification on the original electrocardiosignals according to the majority class characteristics and the minority class characteristics to synthesize samples so as to obtain a classification result. The system comprises: the device comprises a preprocessing module, a feature extraction module, a sample synthesis module and a classification module. By using the method and the device, the depth characteristics of the ECG information can be deeply mined for generating the sample so as to improve the classification accuracy of the model. The method and the system for classifying the unbalanced single-lead electrocardiogram data can be widely applied to the field of signal classification.
Description
Technical Field
The invention relates to the field of signal classification, in particular to a method and a system for classifying unbalanced single-lead electrocardiogram data.
Background
Currently, deep learning methods have been successful in various fields such as target detection, medical image anomaly detection and electrocardiogram intelligent diagnosis, but most of such methods are constructed based on balanced data, and when the methods are faced with real unbalanced data, the methods are severely limited, especially in the medical field, such as the electrocardiogram intelligent auxiliary diagnosis based on deep learning, with the development of wearable equipment and the internet of things, a patient can acquire electrocardiogram signals at any time, and when the quantity ratio of normal electrocardiogram signals to abnormal electrocardiogram signals is seriously unbalanced, the electrocardiogram intelligent auxiliary algorithm is easily misclassified, and a small number of abnormal electrocardiogram signals are easily misclassified into normal electrocardiogram signals.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide an unbalanced single-lead electrocardiogram data classification method and system, which combines a convolutional neural network with a few synthesis oversampling technologies to classify unbalanced electrocardiogram data sets and improve the model classification effect.
The first technical scheme adopted by the invention is as follows: an unbalanced single-lead electrocardiogram data classification method comprises the following steps:
acquiring an original electrocardiosignal and preprocessing the original electrocardiosignal to obtain a preprocessed electrocardiosignal;
carrying out depth feature extraction on the preprocessed electrocardiosignals and reducing dimensions of the depth features to obtain a majority class of features and a minority class of features;
synthesizing the minority class of features based on the SMOTE algorithm to obtain a minority class of feature synthesis sample;
and performing SVM classification on the original electrocardiosignals according to the majority class characteristics and the minority class characteristics to synthesize samples so as to obtain a classification result.
Further, the step of obtaining an original electrocardiographic signal and preprocessing the original electrocardiographic signal to obtain a preprocessed electrocardiographic signal specifically includes:
acquiring an original electrocardiosignal;
carrying out six-layer decomposition on the original electrocardiosignals based on discrete wavelet transform to obtain high-frequency components and low-frequency components of corresponding layers;
discarding the first layer high-frequency component, the second layer high-frequency component and the sixth layer low-frequency component, and reconstructing the remaining high-frequency component and low-frequency component to obtain a reconstructed electrocardiosignal;
and cutting and zero padding are carried out on the reconstructed electrocardiosignals, and the length of the reconstructed electrocardiosignal data is controlled to be 30s, so that the preprocessed electrocardiosignals are obtained.
Further, the step of performing depth feature extraction on the preprocessed electrocardiosignals and performing dimension reduction on the depth features to obtain a majority class of features and a minority class of features specifically includes:
performing depth feature extraction on the preprocessed electrocardiosignals based on a pre-constructed convolutional neural network to obtain 64-dimensional features corresponding to a plurality of samples and a plurality of samples;
and performing dimensionality reduction processing on 64-dimensional features corresponding to the majority samples and the minority samples based on a PCA dimensionality reduction algorithm to obtain 41-dimensional majority features and minority features.
Further, the pre-constructed convolutional neural network comprises four convolutional layers, three maximum pooling layers and three full-connected layers.
Further, the step of synthesizing a minority class of features based on the SMOTE algorithm to obtain a minority class of feature synthesis sample specifically includes:
for each 41-dimensional minority feature X, calculating the distance from X to all other features by taking the Euclidean distance as a standard to obtain corresponding k neighbors;
setting a sampling ratio and determining a sampling multiple N according to the unbalanced ratio of the majority class features and the minority class features;
for each minority class of features X, randomly selecting a plurality of features from the corresponding k neighbors according to the sampling multiple N, and selecting neighbor Xn;
and reconstructing the selected neighbor Xn to obtain a few class characteristic synthetic samples.
Further, the formula for reconstructing the sample for the selected neighbor Xn is as follows:
in the above formula, rand (0,1) represents randomly selecting an array from 0-1, X represents a few kinds of features,representing the selected K neighbors.
The second technical scheme adopted by the invention is as follows: an unbalanced single lead electrocardiographic data classification system comprising:
the preprocessing module is used for acquiring an original electrocardiosignal and preprocessing the original electrocardiosignal to obtain a preprocessed electrocardiosignal;
the feature extraction module is used for carrying out depth feature extraction on the preprocessed electrocardiosignals and reducing dimensions of the depth features to obtain a majority of features and a minority of features;
the sample synthesis module is used for synthesizing the minority class of features based on the SMOTE algorithm to obtain a minority class of feature synthesis sample;
and the classification module is used for carrying out SVM classification on the original electrocardiosignals according to the majority class characteristics and the minority class characteristics to synthesize samples so as to obtain a classification result.
The method and the system have the beneficial effects that: according to the method, depth feature calculation based on a convolutional neural network architecture is combined with a SMOTE algorithm for generating a few types of samples, and the depth features capable of deeply mining ECG information are subjected to sample generation so as to improve the classification accuracy of the model.
Drawings
FIG. 1 is a flow chart of the steps of a method for classifying unbalanced single-lead electrocardiographic data according to the present invention;
FIG. 2 is a data processing diagram according to an embodiment of the present invention;
FIG. 3 is a block diagram of an unbalanced single lead classification system for ECG data;
FIG. 4 is a schematic exploded view of an ECG signal according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a convolutional neural network according to an embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the figures and the specific embodiments. The step numbers in the following embodiments are provided only for convenience of illustration, the order between the steps is not limited at all, and the execution order of each step in the embodiments can be adapted according to the understanding of those skilled in the art.
Referring to fig. 1 and 2, the present invention provides an unbalanced single-lead electrocardiographic data classification method, which comprises the following steps:
acquiring an original electrocardiosignal and preprocessing the original electrocardiosignal to obtain a preprocessed electrocardiosignal;
carrying out depth feature extraction on the preprocessed electrocardiosignals and reducing dimensions of the depth features to obtain a majority class of features and a minority class of features;
synthesizing the minority class of features based on the SMOTE algorithm to obtain a minority class of feature synthesis sample;
and performing SVM classification on the original electrocardiosignals according to the majority class characteristics and the minority class characteristics to synthesize samples so as to obtain a classification result.
Specifically, the SVM is used for replacing a SoftMax classifier in the CNN, and the classification effect of the training model is improved.
As a preferred embodiment of the method, the step of obtaining the original electrocardiographic signal and preprocessing the original electrocardiographic signal to obtain a preprocessed electrocardiographic signal specifically includes:
acquiring an original electrocardiosignal;
carrying out six-layer decomposition on the original electrocardiosignals based on discrete wavelet transform to obtain high-frequency components and low-frequency components of corresponding layers;
specifically, the original electrocardiographic signal is decomposed in multiple layers by discrete wavelet transform, each layer includes a high-frequency component and a low-frequency component, and the high-frequency component and the low-frequency component of the next layer are decomposed from the low-frequency component of the previous layer, and the exploded view is shown in fig. 4.
Discarding the first layer high-frequency component, the second layer high-frequency component and the sixth layer low-frequency component, and reconstructing the remaining high-frequency component and low-frequency component to obtain a reconstructed electrocardiosignal;
and cutting and zero padding are carried out on the reconstructed electrocardiosignals, and the length of the reconstructed electrocardiosignal data is controlled to be 30s, so that the preprocessed electrocardiosignals are obtained.
Specifically, considering that the convolutional neural network requires equal length of input data of training data and prediction data, the ECG in the data set needs to be cut and zero-filled, the data with the time length longer than 30s is cut, only the last 30s data is retained, and zero-filling is performed on the data with the time length shorter than 30s, so that the data length is 30 s. After the clipping and zero padding operations, all data lengths are 30 s.
Further, as a preferred embodiment of the method, the step of performing depth feature extraction on the preprocessed electrocardiographic signals and performing dimension reduction on the depth features to obtain a majority of features and a minority of features specifically includes:
performing depth feature extraction on the preprocessed electrocardiosignals based on a pre-constructed convolutional neural network to obtain 64-dimensional features corresponding to a plurality of samples and a plurality of samples;
and performing dimensionality reduction processing on 64-dimensional features corresponding to the majority samples and the minority samples based on a PCA dimensionality reduction algorithm to obtain 41-dimensional majority features and minority features.
Further as a preferred embodiment of the method, the pre-constructed convolutional neural network comprises four convolutional layers, three max-pooling layers and three full-connected layers.
Specifically, a schematic diagram of the pre-constructed convolutional neural network structure is shown in fig. 5.
Further, as a preferred embodiment of the method, the step of synthesizing the minority class of features based on the SMOTE algorithm to obtain a minority class feature synthesis sample specifically includes:
for each 41-dimensional minority feature X, calculating the distance from X to all other features by taking the Euclidean distance as a standard to obtain corresponding k neighbors;
setting a sampling ratio and determining a sampling multiple N according to the unbalanced ratio of the majority class features and the minority class features;
for each minority class of features X, randomly selecting a plurality of features from the corresponding k neighbors according to the sampling multiple N, and selecting neighbor Xn;
and reconstructing the selected neighbor Xn to obtain a few class characteristic synthetic samples.
Further as a preferred embodiment of the method, the formula for reconstructing the sample for the selected neighbor Xn is as follows:
in the above formula, rand (0,1) represents randomly selecting an array from 0-1, X represents a few kinds of features,representing the selected K neighbors.
As shown in fig. 3, an unbalanced single lead electrocardiographic data classification system comprises:
the geological disaster position map module is used for acquiring the positions of the hidden danger points of the geological disaster and drawing a geological disaster position map;
the preprocessing module is used for acquiring an original electrocardiosignal and preprocessing the original electrocardiosignal to obtain a preprocessed electrocardiosignal;
the feature extraction module is used for carrying out depth feature extraction on the preprocessed electrocardiosignals and reducing dimensions of the depth features to obtain a majority of features and a minority of features;
the sample synthesis module is used for synthesizing the minority class of features based on the SMOTE algorithm to obtain a minority class of feature synthesis sample;
and the classification module is used for carrying out SVM classification on the original electrocardiosignals according to the majority class characteristics and the minority class characteristics to synthesize samples so as to obtain a classification result.
The contents in the above method embodiments are all applicable to the present system embodiment, the functions specifically implemented by the present system embodiment are the same as those in the above method embodiment, and the beneficial effects achieved by the present system embodiment are also the same as those achieved by the above method embodiment.
While the preferred embodiments of the present invention have been illustrated and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (7)
1. An unbalanced single-lead electrocardiogram data classification method is characterized by comprising the following steps:
acquiring an original electrocardiosignal and preprocessing the original electrocardiosignal to obtain a preprocessed electrocardiosignal;
carrying out depth feature extraction on the preprocessed electrocardiosignals and reducing dimensions of the depth features to obtain a majority class of features and a minority class of features;
synthesizing the minority class of features based on the SMOTE algorithm to obtain a minority class of feature synthesis sample;
and performing SVM classification on the original electrocardiosignals according to the majority class characteristics and the minority class characteristics to synthesize samples so as to obtain a classification result.
2. The method for classifying unbalanced single-lead electrocardiographic data according to claim 1, wherein the step of obtaining and pre-processing the original electrocardiographic signals to obtain pre-processed electrocardiographic signals specifically comprises:
acquiring an original electrocardiosignal;
carrying out six-layer decomposition on the original electrocardiosignals based on discrete wavelet transform to obtain high-frequency components and low-frequency components of corresponding layers;
discarding the first layer high-frequency component, the second layer high-frequency component and the sixth layer low-frequency component, and reconstructing the remaining high-frequency component and low-frequency component to obtain a reconstructed electrocardiosignal;
and cutting and zero padding are carried out on the reconstructed electrocardiosignals, and the length of the reconstructed electrocardiosignal data is controlled to be 30s, so that the preprocessed electrocardiosignals are obtained.
3. The method for classifying unbalanced single-lead electrocardiographic data according to claim 2, wherein the step of performing depth feature extraction on the preprocessed electrocardiographic signals and performing dimension reduction on the depth features to obtain a majority class feature and a minority class feature specifically comprises:
performing depth feature extraction on the preprocessed electrocardiosignals based on a pre-constructed convolutional neural network to obtain 64-dimensional features corresponding to a plurality of samples and a plurality of samples;
and performing dimensionality reduction processing on 64-dimensional features corresponding to the majority samples and the minority samples based on a PCA dimensionality reduction algorithm to obtain 41-dimensional majority features and minority features.
4. The method of classifying unbalanced single-lead electrocardiographic data according to claim 3 wherein the pre-constructed convolutional neural network comprises four convolutional layers, three max-pooling layers and three full-connected layers.
5. The method for classifying unbalanced single-lead electrocardiographic data according to claim 4, wherein the step of synthesizing the minority class features based on the SMOTE algorithm to obtain the minority class feature synthesis samples specifically comprises:
for each 41-dimensional minority feature X, calculating the distance from X to all other features by taking the Euclidean distance as a standard to obtain corresponding k neighbors;
setting a sampling ratio and determining a sampling multiple N according to the unbalanced ratio of the majority class features and the minority class features;
for each minority class of features X, randomly selecting a plurality of features from the corresponding k neighbors according to the sampling multiple N, and selecting neighbor Xn;
and reconstructing the selected neighbor Xn to obtain a few class characteristic synthetic samples.
6. The method of classifying unbalanced single lead electrocardiographic data according to claim 5 wherein the formula for reconstructing samples from the selected neighbor Xn is as follows:
7. An unbalanced single lead electrocardiographic data classification system comprising:
the preprocessing module is used for acquiring an original electrocardiosignal and preprocessing the original electrocardiosignal to obtain a preprocessed electrocardiosignal;
the feature extraction module is used for carrying out depth feature extraction on the preprocessed electrocardiosignals and reducing dimensions of the depth features to obtain a majority of features and a minority of features;
the sample synthesis module is used for synthesizing the minority class of features based on the SMOTE algorithm to obtain a minority class of feature synthesis sample;
and the classification module is used for carrying out SVM classification on the original electrocardiosignals according to the majority class characteristics and the minority class characteristics to synthesize samples so as to obtain a classification result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110258782.9A CN112932497A (en) | 2021-03-10 | 2021-03-10 | Unbalanced single-lead electrocardiogram data classification method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110258782.9A CN112932497A (en) | 2021-03-10 | 2021-03-10 | Unbalanced single-lead electrocardiogram data classification method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112932497A true CN112932497A (en) | 2021-06-11 |
Family
ID=76229084
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110258782.9A Pending CN112932497A (en) | 2021-03-10 | 2021-03-10 | Unbalanced single-lead electrocardiogram data classification method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112932497A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113434401A (en) * | 2021-06-24 | 2021-09-24 | 杭州电子科技大学 | Software defect prediction method based on sample distribution characteristics and SPY algorithm |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107301409A (en) * | 2017-07-18 | 2017-10-27 | 云南大学 | Learn the system and method for processing electrocardiogram based on Wrapper feature selectings Bagging |
CN108113666A (en) * | 2017-12-19 | 2018-06-05 | 中国科学院深圳先进技术研究院 | Recognition methods, device and the equipment of atrial fibrillation signal |
CN109961017A (en) * | 2019-02-26 | 2019-07-02 | 杭州电子科技大学 | A kind of cardiechema signals classification method based on convolution loop neural network |
CN110226920A (en) * | 2019-06-26 | 2019-09-13 | 广州视源电子科技股份有限公司 | Electrocardiosignal recognition methods, device, computer equipment and storage medium |
US20200069205A1 (en) * | 2018-08-30 | 2020-03-05 | Tata Consultancy Services Limited | Non-invasive detection of coronary heart disease from short single-lead ecg |
CN110991653A (en) * | 2019-12-10 | 2020-04-10 | 电子科技大学 | Method for classifying unbalanced data sets |
CN111178399A (en) * | 2019-12-13 | 2020-05-19 | 腾讯科技(深圳)有限公司 | Data processing method and device, electronic equipment and computer readable storage medium |
CN111202512A (en) * | 2020-03-05 | 2020-05-29 | 齐鲁工业大学 | Electrocardiogram classification method and device based on wavelet transformation and DCNN |
-
2021
- 2021-03-10 CN CN202110258782.9A patent/CN112932497A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107301409A (en) * | 2017-07-18 | 2017-10-27 | 云南大学 | Learn the system and method for processing electrocardiogram based on Wrapper feature selectings Bagging |
CN108113666A (en) * | 2017-12-19 | 2018-06-05 | 中国科学院深圳先进技术研究院 | Recognition methods, device and the equipment of atrial fibrillation signal |
US20200069205A1 (en) * | 2018-08-30 | 2020-03-05 | Tata Consultancy Services Limited | Non-invasive detection of coronary heart disease from short single-lead ecg |
CN109961017A (en) * | 2019-02-26 | 2019-07-02 | 杭州电子科技大学 | A kind of cardiechema signals classification method based on convolution loop neural network |
CN110226920A (en) * | 2019-06-26 | 2019-09-13 | 广州视源电子科技股份有限公司 | Electrocardiosignal recognition methods, device, computer equipment and storage medium |
CN110991653A (en) * | 2019-12-10 | 2020-04-10 | 电子科技大学 | Method for classifying unbalanced data sets |
CN111178399A (en) * | 2019-12-13 | 2020-05-19 | 腾讯科技(深圳)有限公司 | Data processing method and device, electronic equipment and computer readable storage medium |
CN111202512A (en) * | 2020-03-05 | 2020-05-29 | 齐鲁工业大学 | Electrocardiogram classification method and device based on wavelet transformation and DCNN |
Non-Patent Citations (1)
Title |
---|
候弘慧: ""基于特征融合的不平衡ECG信号分析"", 中国优秀硕士学位论文全文数据库(信息科技辑), no. 4, 15 April 2019 (2019-04-15), pages 1 - 6 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113434401A (en) * | 2021-06-24 | 2021-09-24 | 杭州电子科技大学 | Software defect prediction method based on sample distribution characteristics and SPY algorithm |
CN113434401B (en) * | 2021-06-24 | 2022-10-28 | 杭州电子科技大学 | Software defect prediction method based on sample distribution characteristics and SPY algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Salama et al. | EEG-based emotion recognition using 3D convolutional neural networks | |
Chaovalit et al. | Discrete wavelet transform-based time series analysis and mining | |
Batal et al. | A supervised time series feature extraction technique using dct and dwt | |
CN112766355A (en) | Electroencephalogram signal emotion recognition method under label noise | |
CN111310656A (en) | Single motor imagery electroencephalogram signal identification method based on multi-linear principal component analysis | |
CN114176607B (en) | Electroencephalogram signal classification method based on vision transducer | |
CN111898526B (en) | Myoelectric gesture recognition method based on multi-stream convolution neural network | |
CN111639697B (en) | Hyperspectral image classification method based on non-repeated sampling and prototype network | |
CN106599903B (en) | Signal reconstruction method for weighted least square dictionary learning based on correlation | |
Zhu et al. | TCRAN: Multivariate time series classification using residual channel attention networks with time correction | |
CN117009780A (en) | Space-time frequency domain effective channel attention motor imagery brain electrolysis code method based on contrast learning | |
CN112932497A (en) | Unbalanced single-lead electrocardiogram data classification method and system | |
CN108537123A (en) | Electrocardiogram recognition method based on multi-feature extraction | |
CN115530788A (en) | Arrhythmia classification method based on self-attention mechanism | |
CN115204231A (en) | Digital human-computer interface cognitive load assessment method based on EEG (electroencephalogram) multi-dimensional features | |
Wang et al. | Representation learning with deconvolution for multivariate time series classification and visualization | |
Maharaj et al. | Discrimination of locally stationary time series using wavelets | |
Kwon et al. | Accurate blind Lempel-Ziv-77 parameter estimation via 1-D to 2-D data conversion over convolutional neural network | |
CN112932503B (en) | Motor imagery task decoding method based on 4D data expression and 3DCNN | |
CN114139572A (en) | Electroencephalogram emotion recognition method based on enhanced symmetric positive definite matrix | |
CN112800928A (en) | Epileptic seizure prediction method of global self-attention residual error network with channel and spectrum features fused | |
CN115844424B (en) | Sleep spindle wave hierarchical identification method and system | |
CN113869289B (en) | Multi-channel ship radiation noise feature extraction method based on entropy | |
Wilson et al. | Deep riemannian networks for eeg decoding | |
CN112560784B (en) | Electrocardiogram classification method based on dynamic multi-scale convolutional neural network |
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 |