The performance of traditional classification algorithms can be limited on imbalanced datasets. In recent years, the imbalanced data learning problem has drawn significant interest. In this work, we focus on designing...
详细信息
The performance of traditional classification algorithms can be limited on imbalanced datasets. In recent years, the imbalanced data learning problem has drawn significant interest. In this work, we focus on designing modifications to neural network, in order to appropriately tackle the problem of multiclass imbalance. We propose a hybrid method that combines two ideas: diverse random subspace ensemble learning with evolutionary search, to improve the performance of neural network on multiclass imbalanced data. An evolutionary search technique is utilized to optimize the misclassification cost under the guidance of imbalanced data measures. Moreover, the diverse random subspace ensemble employs the minimum overlapping mechanism to provide diversity so as to improve the performance of the learning and optimization of neural network. We have demonstrated experimentally using UCI datasets that our approach can achieve better result than state-of-the-art methods for imbalanced data.
Aiming at avoiding misregistration in complicated medicalimage registration based on SURF (Speed-Up Robust Features)-TPS (Thin-Plate Spline), we propose a novel algorithm. This method is based on SURF and human inter...
详细信息
Traditional medicalimage enhancement method has some disadvantages. They can not significantly improve the medicalimage edge, texture and detailed information. Besides the enhancement effect is susceptible to interf...
详细信息
Breast tumor detection is a most effective way to immunized against mammary *** is known that the sort algorithm of extreme learning machine(ELM),in view of the feature model for breast X-ray image,is being applied in...
详细信息
Breast tumor detection is a most effective way to immunized against mammary *** is known that the sort algorithm of extreme learning machine(ELM),in view of the feature model for breast X-ray image,is being applied in the computer aided detection of breast *** the basis of all these,it is raised in this paper that marking for the suspicious region in the double view mammography by the use of ELM,then classifying the result of double views marking by using the Simple Bias classifier and finally gaining the detection *** experiment with 444 cases or 222 pair of X-ray mammography from Liao Ning Province Cancer Hospital shows that,the breast tumor detection in double views mammography based on Simple Bias is an available and effective way to detect breast tumor.
Many lung nodule computer-aided detection methods have been proposed to help radiologists in their decision making. Because high sensitivity is essential in the candidate identification stage, there are countless fals...
详细信息
ISBN:
(纸本)9781479910526
Many lung nodule computer-aided detection methods have been proposed to help radiologists in their decision making. Because high sensitivity is essential in the candidate identification stage, there are countless false positives produced by the initial suspect nodule generation process, giving more work to radiologists. The difficulty of false positive reduction lies in the variation of the appearances of the potential nodules, and the imbalance distribution between the amount of nodule and non-nodule candidates in the dataset. To solve these challenges, we extend the random subspace method to a novel Cost Sensitive Adaptive Random Subspace ensemble (CSARS), so as to increase the diversity among the components and overcome imbalanced data classification. Experimental results show the effectiveness of the proposed method in terms of G-mean and AUC in comparison with commonly used methods.
Internet worm can spread not only via networks but also through external removable *** this paper,considering the influence of removable media,we construct two new worm propagation *** analyzing the local asymptotic s...
详细信息
Internet worm can spread not only via networks but also through external removable *** this paper,considering the influence of removable media,we construct two new worm propagation *** analyzing the local asymptotic stability at infection-free equilibrium,a basic reproduction number is *** is locally stable at infection-free equilibrium if basic reproduction number is less than one;whereas system will lose its *** numerical experiments are given to support the theoretical ***,simulating the propagation behavior via simulation experiments,we further demonstrate the correctness of theoretical analysis.
Abdominal anatomical structure database is a necessary precondition for medicalimage computer-aided *** this paper,a Web database system based on the CT abdomen anatomy images is *** system consists of abdominal visc...
详细信息
Abdominal anatomical structure database is a necessary precondition for medicalimage computer-aided *** this paper,a Web database system based on the CT abdomen anatomy images is *** system consists of abdominal viscera segmentation and anatomical structure ***,we need to segment the spleen and liver in abdominal CT *** spleen segmentation,an improved Otsu-based algorithm is put forward;for liver segmentation,an improved algorithm combined probability graph and traditional graph cut segmentation algorithm is ***,a Web oriented abdominal viscera anatomical structure database system using SSI framework is constructed,storing and managing image information of segmented ***,through the experimental measurement for abdominal CT image segmentation and database system,the experimental results show that the proposed segmentation algorithm and the liver spleen segmentation algorithm can get more accurate segmentation *** database system can provide remote management,making it available for medical researchers use the computer remotely access to the Internet to obtain the required data anywhere.1
The method on gene classification has been widely studied with the development of gene *** learning is the best choice to research the *** both traditional SVM and ELM cannot fulfill the requirement of high accuracy a...
详细信息
The method on gene classification has been widely studied with the development of gene *** learning is the best choice to research the *** both traditional SVM and ELM cannot fulfill the requirement of high accuracy and short ***,in this paper,we propose a novel Accuracy Adaptive Extreme Learning Machine(A 2-ELM) which can cover the shortage of traditional SVM and ELM in the fact of more ***,we propose a method of feature selection and overview the property of traditional ***,an Accuracy of Adaptive ELM(A 2-ELM) is developed,which can fulfill the requirement for accurately and ***,we conduct experiments on gene expression data to verify the dynamic and accurate of our proposed accuracy of adaptive ELM in classification gene expression data with experimental settings.
Currently,dental CBCT(Cone-Beam Computed Tomography) is wildly used in the tooth implantation,maxillofacial surgery and the treatment of temporomandibular joint *** three-dimensional images of Dental CBCT are reconstr...
详细信息
Currently,dental CBCT(Cone-Beam Computed Tomography) is wildly used in the tooth implantation,maxillofacial surgery and the treatment of temporomandibular joint *** three-dimensional images of Dental CBCT are reconstructed based on the projection images,which are obtained at each degree of *** projected image should have a precise correspondence with the angle of rotation of *** vibration occurs during the rotation,artifacts will be caused from reconstruction,thus affecting the image *** has a very important significance on the image reconstruction if vibration information of CBCT arm can be detected and *** paper presented a vibration measuring device based 3D accelerometer,which can measure and evaluate vibration of CBCT during *** this device,vibration information can be sensed by 3D *** analog signals are sampled by Microprocessor with internal ADC ***,all data are sent to PC via wireless communication *** written by MATLAB will acquire,display,save and analyze data from measurement device,which makes it possible that evaluates vibration information from CBCT.
3D display can reflect the image depth information and make the image present stereoscopic sensation to improve the medical imaging effect and diagnosis accuracy,if the technology can be used in image guide surgery to...
详细信息
3D display can reflect the image depth information and make the image present stereoscopic sensation to improve the medical imaging effect and diagnosis accuracy,if the technology can be used in image guide surgery to diagnose lesion *** paper will mainly elaborate some advanced naked eye 3D display technologies--multi-view auto-stereoscopic display technology,integral photography 3D display technology and holographic 3D display *** last,image guide surgery based on interactive 3D navigation system is ***,this paper indicates the trend of medical imaging fields.
暂无评论