检索结果-内蒙古大学图书馆

Learning data-adaptive nonparametric kernels

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Liu, Fanghui Huang, Xiaolin Gong, Chen Yang, Jie Li, Li Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai200240 China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information Ministry of Education School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China Department of Automation Tsinghua University

Kernel methods have been extensively used in a variety of machine learning tasks such as classification, clustering, and dimensionality reduction. For complicated practical tasks, the traditional kernels, e.g., Gaussian kernel and sigmoid kernel, or their combinations are often not sufficiently flexible to fit the data. In this paper, we present a Data-Adaptive Nonparametric Kernel (DANK) learning framework in a data-driven manner. To be specific, in model formulation, we impose an adaptive matrix on the kernel/Gram matrix in an entry-wise strategy. Since we do not specify the formulation of the adaptive matrix, each entry in the adaptive matrix can be directly and flexibly learned from the data. Therefore, the solution space of the learned kernel is largely expanded, which makes our DANK model flexible to capture the data with different local statistical properties. Specifically, the proposed kernel learning framework can be seamlessly embedded to support vector machines (SVM) and support vector regression (SVR), which has the capability of enlarging the margin between classes and reducing the model generalization error. Theoretically, we demonstrate that the objective function of our DANK model embedded in SVM/SVR is gradient-Lipschitz continuous. Thereby, the training process for kernel and parameter learning in SVM/SVR can be efficiently optimized in a unified framework. Further, to address the scalability issue in nonparametric kernel learning framework, we decompose the entire optimization problem in DANK into several smaller easy-to-solve problems, so that our DANK model can be efficiently approximated by this partition. The effectiveness of this approximation is demonstrated by both empirical studies and theoretical guarantees. Experimentally, the proposed DANK model embedded in SVM/SVR achieves encouraging performance on various classification and regression benchmark datasets when compared with other representative kernel learning based algorithms. Copyrig

关键词： Matrix algebra

Hierarchical P systems with randomized right-hand sides of rules 18th

学校读者我要写书评

暂无评论

Hierarchical P systems with randomized right-hand sides of r...

18th International Conference on Membrane Computing, CMC 1018

作者： Alhazov, Artiom Freund, Rudolf Ivanov, Sergiu Institute of Mathematics and Computer Science Academy of Sciences of Moldova Academiei 5 Chişinău2028 Moldova Key Laboratory of Image Information Processing and Intelligent Control of Education Ministry of China School of Automation Huazhong University of Science and Technology Wuhan430074 China Faculty of Informatics TU Wien Favoritenstraße 9–11 Vienna1040 Austria LACL Université Paris Est – Créteil Val de Marne 61 av. Général de Gaulle Créteil94010 France TIMC-IMAG/DyCTiM Faculty of Medicine of Grenoble 5 avenue du Grand Sablon La Tronche38700 France

ISBN: (纸本)9783319733586

P systems are a model of hierarchically compartmentalized multiset rewriting. We introduce a novel kind of P systems in which rules are dynamically constructed in each step by non-deterministic pairing of left-hand and right-hand sides. We define three variants of right-hand side randomization and compare each of them with the power of conventional P systems. It turns out that all three variants enable non-cooperative P systems to generate exponential (and thus non-semi-linear) number languages. We also give a binary normal form for one of the variants of P systems with randomized rule right-hand sides. © 2018, Springer International Publishing AG.

关键词： Hierarchical systems

A novel color image encryption algorithm based on genetic recombination and the four-dimensional memristive hyperchaotic system

学校读者我要写书评

暂无评论

Chinese Physics B 2016年第10期25卷 76-88页

作者：柴秀丽甘志华路杨张苗辉陈怡然 School of Computer and Information Engineering Institute of Image Processing and Pattern Recognition Henan University Department of Electrical and Computer Engineering University of Pittsburgh School of Software Henan University Research Department Henan University

Recently, many image encryption algorithms based on chaos have been proposed. Most of the previous algorithms encrypt components R, G, and B of color images independently and neglect the high correlation between them. In the paper, a novel color image encryption algorithm is introduced. The 24 bit planes of components R, G, and B of the color plain image are obtained and recombined into 4 compound bit planes, and this can make the three components affect each other. A four-dimensional(4D) memristive hyperchaotic system generates the pseudorandom key streams and its initial values come from the SHA 256 hash value of the color plain image. The compound bit planes and key streams are confused according to the principles of genetic recombination, then confusion and diffusion as a union are applied to the bit planes,and the color cipher image is obtained. Experimental results and security analyses demonstrate that the proposed algorithm is secure and effective so that it may be adopted for secure communication.

关键词： genetic recombination memristive hyperchaotic system color image encryption confusion diffusion

Generalization Properties of hyper-RKHS and its Applications

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Liu, Fanghui Shi, Lei Huang, Xiaolin Yang, Jie Suykens, Johan A.K. Department of Electrical Engineering ESAT-STADIUS KU Leuven Kasteelpark Arenberg 10 LeuvenB-3001 Belgium Shanghai Key Laboratory for Contemporary Applied Mathematics School of Mathematical Sciences Fudan University Shanghai200433 China Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Institute of Medical Robotics Shanghai Jiao Tong University Shanghai200240 China

This paper generalizes regularized regression problems in a hyper-reproducing kernel Hilbert space (hyper-RKHS), illustrates its utility for kernel learning and out-of-sample extensions, and proves asymptotic convergence results for the introduced regression models in an approximation theory view. Algorithmically, we consider two regularized regression models with bivariate forms in this space, including kernel ridge regression (KRR) and support vector regression (SVR) endowed with hyper-RKHS, and further combine divide-and-conquer with Nyström approximation for scalability in large sample cases. This framework is general: the underlying kernel is learned from a broad class, and can be positive definite or not, which adapts to various requirements in kernel learning. Theoretically, we study the convergence behavior of regularized regression algorithms in hyper-RKHS and derive the learning rates, which goes beyond the classical analysis on RKHS due to the non-trivial independence of pairwise samples and the characterisation of hyper-RKHS. Experimentally, results on several benchmarks suggest that the employed framework is able to learn a general kernel function form an arbitrary similarity matrix, and thus achieves a satisfactory performance on classification tasks. Copyright © 2018, The Authors. All rights reserved.

关键词： Approximation theory

FWLBP: A scale invariant descriptor for texture classification

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Roy, Swalpa Kumar Bhattacharya, Nilavra Chanda, Bhabatosh Chaudhuri, Bidyut B. Ghosh, Dipak Kumar Optical Character Recognition Laboratory Computer Vision and Pattern Recognition Unit Indian Statistical Institute Kolkata700108 India School of Information University of Texas AustinTX78712 United States Image Processing Laboratory Electronics and Communication Sciences Unit Indian Statistical Institute Kolkata700108 India Department of Electronics and Communication Engineering National Institute of Technology Rourkela Rourkela769008 India

In this paper we propose a novel texture descriptor called Fractal Weighted Local Binary Pattern (FWLBP). The fractal dimension (FD) measure is relatively invariant to scale-changes, and presents a good correlation with human viewpoint of surface roughness. We have utilized this property to construct a scale-invariant descriptor. Here, the input image is sampled using an augmented form of the local binary pattern (LBP) over three different radii, and then used an indexing operation to assign FD weights to the collected samples. The final histogram of the descriptor has its features calculated using LBP, and its weights computed from the FD image. The proposed descriptor is scale invariant, and is also robust in rotation or reflection, and partially tolerant to noise and illumination changes. In addition, the local fractal dimension is relatively insensitive to the bi-Lipschitz transformations, whereas its extension is adequate to precisely discriminate the fundamental of texture primitives. Experiment results carried out on standard texture databases show that the proposed descriptor achieved better classification rates compared to the state-of-the-art descriptors. Copyright © 2018, The Authors. All rights reserved.

关键词： Fractal dimension

Visual attention network for low dose CT

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Du, Wenchao Chen, Hu Liao, Peixi Yang, Hongyu Wang, Ge Zhang, Yi College of Computer Science Sichuan University Chengdu610065 China Guangdong Provincial Key Laboratory of Medical Image Processing Southern Medical University Guangzhou510515 China Department of Scientific Research and Education Sixth People’s Hospital of Chengdu Chengdu610065 China Department of Biomedical Engineering Rensselaer Polytechnic Institute TroyNY12180 United States

Noise and artifacts are intrinsic to low dose CT (LDCT) data acquisition, and will significantly affect the imaging performance. Perfect noise removal and image restoration is intractable in the context of LDCT due to the statistical and technical uncertainties. In this paper, we apply the generative adversarial network (GAN) framework with a visual attention mechanism to deal with this problem in a data-driven/machine learning fashion. Our main idea is to inject visual attention knowledge into the learning process of GAN to provide a powerful prior of the noise distribution. By doing this, both the generator and discriminator networks are empowered with visual attention information so they will not only pay special attention to noisy regions and surrounding structures but also explicitly assess the local consistency of the recovered regions. Our experiments qualitatively and quantitatively demonstrate the effectiveness of the proposed method with clinic CT images. Copyright © 2018, The Authors. All rights reserved.

关键词： Generative adversarial networks

Micro EEG/ECG signal's chopper-stabilization amplifying chip for novel drycontact electrode

学校读者我要写书评

暂无评论

Journal of Semiconductors 2017年第2期38卷 96-104页

作者： Jianhui Sun Chunxing Wang Gongtang Wang Jinhui Wang Qing Hua Chuanfu Cheng Xinxia Cai Tao Yin Yang Yu Haigang Yang Dengwang Li College of Physics and Electronics Shandong Normal University Jinan 240014 China Institute of Electronics Chinese Academy of Sciences Beijing 100190 China Shandong Province Key Laboratory of Medical Physics and Image Processing Technology Institute of BioelectronicsShandong Normal University Jinan 240014 China Department of Electronical and Computer Engineering North Dakota State University Fargo ND 58102 USA

Facing the body＇s EEG（electroencephalograph, 0.5–100 Hz, 5–100 μV） and ECG＇s（electrocardiogram,〈 100 Hz, 0.01–5 mV） micro signal detection requirement, this paper develops a pervasive application micro signal detection ASIC chip with the chopping modulation/demodulation method. The chopper-stabilization circuit with the RRL（ripple reduction loop） circuit is to suppress the ripple voltage, which locates at the single-stage amplifier＇s outputting terminal. The single-stage chopping core＇s noise has been suppressed too, and it is beneficial for suppressing noises of post-circuit. The chopping core circuit uses the PFB（positive feedback loop） to increase the inputting resistance, and the NFB（negative feedback loop） to stabilize the 40 dB intermediate frequency gain. The cascaded switch-capacitor sample/hold circuit has been used for deleting spike noises caused by non-ideal MOS switches, and the VGA/BPF（voltage gain amplifier/band pass filter） circuit is used to tune the chopper system＇s gain/bandwidth digitally. Assisted with the designed novel dry-electrode, the real test result of the chopping amplifying circuit gives some critical parameters： 8.1 μW/channel, 0.8 μVrms（@band-widthD100 Hz）, 4216–11220 times digitally tuning gain range, etc. The data capture system uses the NI CO＇s data capturing DAQmx interface,and the captured micro EEG/ECG＇s waves are real-time displayed with the PC-Labview. The proposed chopper system is a unified EEG/ECG signal＇s detection instrument and has a critical real application value.

关键词： EEG/ECG novel dry-contact electrode weak and micro signal detection chopping modulation/demodulation de-noising gain/band width digitally tuning

The spectrum entropy estimation method for ultrasonic backscattering signal in cancellous bones

学校读者我要写书评

暂无评论

Chinese Journal of Acoustics 2017年第3期36卷 337-346页

作者： XU Feng XU Yiwen LIU Chengcheng TA De'an GUO Jianzhong Department of Electronic Engineering Fudan UniversityShanghai 200433 Institute of Acoustics Tongji UniversityShanghai 200092 Shanghai Key Laboratory of Medical Image Processing and Computer Assisted Surgery Fudan UniversityShanghai 200433 School of Physics Shaanxi Normal UniversityXi'an 710119

In the application of cancellous bone ultrasound diagnosis based on backscattering method, it is of great importance to estimate fast and accurately whether the valid backscattering signal exists in the received signal. We propose a fast estimation method based on spectrum entropy method. With 984 records of adult calcaneus clinical data, we estimate the validity of the backscatter signal using this method. The results of the proposed method and the results of experience-base judgement were compared and analyzed. And two key parameters, the signal range length and the segment number of the spectrum entropy, were analyzed. The results show when the signal range length is 13 I^s and the segment number is 15 20, this method can get the best result （accuracy〉95%, sensitivity〉99%, specificity〉87%）, while taking little calculation time （1.5 ms）. Therefore, this spectrum entropy method can satisfy the accuracy and real-time requirements in the ultrasonic estimation for cancellous bone.

关键词： bone The spectrum entropy estimation method for ultrasonic backscattering signal in cancellous bones

Unsupervised extraction of phenotypes from cancer clinical notes for association studies

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Stark, Stefan G. Hyland, Stephanie L. Pradier, Melanie F. Lehmann, Kjong-Van Wicki, Andreas Perez-Cruz, Fernando Vogt, Julia E. Rätsch, Gunnar Computational Biology Program Memorial Sloan Kettering Cancer Center New York United States Tri-Institutional Ph.D. Program in Computational Biology and Medicine Weill Cornell Medicine New York United States Department of Computer Science ETH Zürich Zürich Switzerland Medical Informatics Group University Hospital Zürich Zürich Switzerland Swiss Institute for Bioinformatics Zurich Switzerland Department of Biology ETH Zürich Zürich Switzerland Department of Signal Processing and Information Theory University Carlos III in Madrid Leganés Spain School of Engineering and Applied Sciences Harvard University CambridgeMA United States Department of Biomedicine University of Basel Basel Switzerland Tumorzentrum University Hospital Basel Basel Switzerland Swiss Data Science Center ETH Zürich and EPFL Lausanne Switzerland Department of Mathematics and Computer Science University of Basel Basel Switzerland

The recent adoption of Electronic Health Records (EHRs) by healthcare providers has introduced an important source of data that provides detailed and highly specific insights into patient phenotypes over large cohorts. These datasets, in combination with machine learning and statistical approaches, generate new opportunities for research and clinical care. However, many methods require the patient representations to be in structured formats, while the information in the EHR is often locked in unstructured text designed for human readability. In this work, we develop the methodology to automatically extract clinical features from clinical narratives from large EHR corpora without the need for prior knowledge. We consider medical terms and sentences appearing in clinical narratives as atomic information units. We propose an efficient clustering strategy suitable for the analysis of large text corpora and utilize the clusters to represent information about the patient compactly. Additionally, we define the sentences on ontologic and natural language vocabularies to automatically detect pertinent combinations of concepts present in the corpus, even when an ontology is not available. To demonstrate the utility of our approach, we perform an association study of clinical features with somatic mutation profiles from 4,007 cancer patients and their tumors. We apply the proposed algorithm to a dataset consisting of .65 thousand documents with a total of .3.2 million sentences. After correcting for cancer type and other confounding factors, we identify a total of 340 significant statistical associations between the presence of somatic mutations and clinical features. We annotated these associations according to their novelty and we report several known associations. We also propose 37 plausible, testable hypothesis for associations where the underlying biological mechanism does not appear to be known. These results illustrate that the automated discovery of clinical features

关键词： Diseases