With the improvement of radar resolution,the dimension of the high resolution range profile(HRRP)has increased.In order to solve the small sample problem caused by the increase of HRRP dimension,an algorithm based on ...With the improvement of radar resolution,the dimension of the high resolution range profile(HRRP)has increased.In order to solve the small sample problem caused by the increase of HRRP dimension,an algorithm based on kernel joint discriminant analysis(KJDA)is proposed.Compared with the traditional feature extraction methods,KJDA possesses stronger discriminative ability in the kernel feature space.K-nearest neighbor(KNN)and kernel support vector machine(KSVM)are applied as feature classifiers to verify the classification effect.Experimental results on the measured aircraft datasets show that KJDA can reduce the dimensionality,and improve target recognition performance.展开更多
Several industrial coal processes are largely determined by the distribution of particle sizes in their feed.Currently these parameters are measured by manual sampling,which is time consuming and cannot provide real t...Several industrial coal processes are largely determined by the distribution of particle sizes in their feed.Currently these parameters are measured by manual sampling,which is time consuming and cannot provide real time feedback for automatic control purposes.In this paper,an approach using image segmentation on images of overlapped coal particles is described.The estimation of the particle size distribution by number is also described.The particle overlap problem was solved using image enhancement algorithms that converted those image parts representing material in lower layers to black.Exponential high-pass filter(EHPF) algorithms were used to remove the texture from particles on the surface.Finally,the edges of the surface particles were identified by morphological edge detection.These algorithms are described in detail as is the method of extracting the coal particle size.Tests indicate that using more coal images gives a higher accuracy estimate.The positive absolute error of 50 random tests was consistently less than 2.5% and the errors were reduced as the size of the fraction increased.展开更多
Zero-shot learning enables the recognition of new class samples by migrating models learned from semanticfeatures and existing sample features to things that have never been seen before. The problems of consistencyof ...Zero-shot learning enables the recognition of new class samples by migrating models learned from semanticfeatures and existing sample features to things that have never been seen before. The problems of consistencyof different types of features and domain shift problems are two of the critical issues in zero-shot learning. Toaddress both of these issues, this paper proposes a new modeling structure. The traditional approach mappedsemantic features and visual features into the same feature space;based on this, a dual discriminator approachis used in the proposed model. This dual discriminator approach can further enhance the consistency betweensemantic and visual features. At the same time, this approach can also align unseen class semantic features andtraining set samples, providing a portion of information about the unseen classes. In addition, a new feature fusionmethod is proposed in the model. This method is equivalent to adding perturbation to the seen class features,which can reduce the degree to which the classification results in the model are biased towards the seen classes.At the same time, this feature fusion method can provide part of the information of the unseen classes, improvingits classification accuracy in generalized zero-shot learning and reducing domain bias. The proposed method isvalidated and compared with othermethods on four datasets, and fromthe experimental results, it can be seen thatthe method proposed in this paper achieves promising results.展开更多
In many practical classification problems,datasets would have a portion of outliers,which could greatly affect the performance of the constructed models.In order to address this issue,we apply the group method of data...In many practical classification problems,datasets would have a portion of outliers,which could greatly affect the performance of the constructed models.In order to address this issue,we apply the group method of data handin neural network in outlier detection.This study builds a GMDH-based outlier detectio model.This model first implements feature selection in the training set L using GMDH neural network.Then a new training set L can be obtained by mapping the selected key feature subset.Next,a linear regression model can be constructed in the set L by ordinary least squares estimation.Further,it eliminates a sample from the set L randomly every time,and then rebuilds a linear regression model.Finally,outlier detection is realized by calculating Cook’s distance for each sample.Four different customer classification datasets are used to conduct experiments.Results show that GOD model can effectively eliminate outliers,and compared with the five existing outlier detection models,it generally performs significantly better.This indicates that eliminating outliers can effectively enhance classification accuracy of the trained classification model.展开更多
基金supported by the National Natural Science Foundation of China(61471191)the Aeronautical Science Foundation of China(20152052026)the Foundation of Graduate Innovation Center in NUAA(kfjj20170313)
文摘With the improvement of radar resolution,the dimension of the high resolution range profile(HRRP)has increased.In order to solve the small sample problem caused by the increase of HRRP dimension,an algorithm based on kernel joint discriminant analysis(KJDA)is proposed.Compared with the traditional feature extraction methods,KJDA possesses stronger discriminative ability in the kernel feature space.K-nearest neighbor(KNN)and kernel support vector machine(KSVM)are applied as feature classifiers to verify the classification effect.Experimental results on the measured aircraft datasets show that KJDA can reduce the dimensionality,and improve target recognition performance.
基金the Creative Research Groups Science Fund of the National Natural Science Foundation of China(No.50921002)
文摘Several industrial coal processes are largely determined by the distribution of particle sizes in their feed.Currently these parameters are measured by manual sampling,which is time consuming and cannot provide real time feedback for automatic control purposes.In this paper,an approach using image segmentation on images of overlapped coal particles is described.The estimation of the particle size distribution by number is also described.The particle overlap problem was solved using image enhancement algorithms that converted those image parts representing material in lower layers to black.Exponential high-pass filter(EHPF) algorithms were used to remove the texture from particles on the surface.Finally,the edges of the surface particles were identified by morphological edge detection.These algorithms are described in detail as is the method of extracting the coal particle size.Tests indicate that using more coal images gives a higher accuracy estimate.The positive absolute error of 50 random tests was consistently less than 2.5% and the errors were reduced as the size of the fraction increased.
文摘Zero-shot learning enables the recognition of new class samples by migrating models learned from semanticfeatures and existing sample features to things that have never been seen before. The problems of consistencyof different types of features and domain shift problems are two of the critical issues in zero-shot learning. Toaddress both of these issues, this paper proposes a new modeling structure. The traditional approach mappedsemantic features and visual features into the same feature space;based on this, a dual discriminator approachis used in the proposed model. This dual discriminator approach can further enhance the consistency betweensemantic and visual features. At the same time, this approach can also align unseen class semantic features andtraining set samples, providing a portion of information about the unseen classes. In addition, a new feature fusionmethod is proposed in the model. This method is equivalent to adding perturbation to the seen class features,which can reduce the degree to which the classification results in the model are biased towards the seen classes.At the same time, this feature fusion method can provide part of the information of the unseen classes, improvingits classification accuracy in generalized zero-shot learning and reducing domain bias. The proposed method isvalidated and compared with othermethods on four datasets, and fromthe experimental results, it can be seen thatthe method proposed in this paper achieves promising results.
基金partly supported by the Major Project of the National Social Science Foundation of China under Grant No.18VZL006the National Natural Science Foundation of China under Grant Nos.71571126and 71974139+6 种基金the Excellent Youth Foundation of Sichuan Province under Grant No.20JCQN0225the Tianfu Ten-thousand Talents Program of Sichuan Provincethe Excellent Youth Foundation of Sichuan University under Grant No.sksyl201709the Leading Cultivation Talents Program of Sichuan Universitythe Teacher and Student Joint Innovation Project of Business School of Sichuan University under Grant No.LH2018011the2018 Special Project for Cultivation and Innovation of New AcademicQian Platform Talent under Grant No.5772-012。
文摘In many practical classification problems,datasets would have a portion of outliers,which could greatly affect the performance of the constructed models.In order to address this issue,we apply the group method of data handin neural network in outlier detection.This study builds a GMDH-based outlier detectio model.This model first implements feature selection in the training set L using GMDH neural network.Then a new training set L can be obtained by mapping the selected key feature subset.Next,a linear regression model can be constructed in the set L by ordinary least squares estimation.Further,it eliminates a sample from the set L randomly every time,and then rebuilds a linear regression model.Finally,outlier detection is realized by calculating Cook’s distance for each sample.Four different customer classification datasets are used to conduct experiments.Results show that GOD model can effectively eliminate outliers,and compared with the five existing outlier detection models,it generally performs significantly better.This indicates that eliminating outliers can effectively enhance classification accuracy of the trained classification model.
