基于拉普拉斯分值和超球支持向量机的轴承故障诊断方法设计

针对轴承振动信号中的故障信息往往很微弱,同时振动样本数据分布不平衡即故障样本占总样本数的比例低,从而导致故障诊断模型训练不精确而影响诊断精度的问题,提出了一种基于拉普拉斯分值和超球大间隔支持向量机的故障诊断方法;首先,采用有标签的训练样本数据和拉普拉斯分值法提取原始振动信号中的微弱故障信息,并降低其数据维数,从而得到用于故障诊断的特征向量,然后设计了一种改进的超球大间隔支持向量机的故障诊断模型,通过最小化超球体积和最大化超球边界和故障样本之间的间隔来实现故障诊断,以解决样本的不均衡问题,最终通过将测试样本数据代入决策方程并通过投票机制确定其故障类别;在Matlab环境下对轴承故障诊断进行实验,实验结果证明了文中方法能有效解决样本的不均衡情况下的故障诊断,且相对其它方法,具有诊断精度高和收敛速度快的优点.     

Analytic center cutting plane method for multiple kernel learning

Multiple Kernel Learning (MKL) is a popular generalization of kernel methods which allows the practitioner to optimize over convex combinations of kernels. We observe that many recent MKL solutions can be cast in the framework of oracle based optimization, and show that they vary in terms of query point generation. The popularity of such methods is because the oracle can fortuitously be implemented as a support vector machine. Motivated by the success of centering approaches in interior point methods, we propose a new approach to optimize the MKL objective based on the analytic center cutting plane method (accpm). Our experimental results show that accpm outperforms state of the art in terms of rate of convergence and robustness. Further analysis sheds some light as to why MKL may not always improve classification accuracy over naive solutions.     

Extraction of enhanced evoked potentials using wavelet filtering

In order to remove physiological artefacts and gain the improved evoked potentials, we propose a filtering method using the multi-resolution wavelet transform. The wavelet transform is repeatedly performed until all resolution levels are obtained. It decomposes the measured evoked potentials into scale coefficients corresponding to low frequency components and wavelet coefficients corresponding to high frequency components. In the wavelet domain, artefacts are dispersed mainly at the wavelet coefficients rather than the scaling coefficients. Thus, when the inverse wavelet transform is performed, this method shrinks the wavelet coefficients to reduce artefacts with shrinkage functions. By repeatedly performing the inverse wavelet transform, an evoked potential having the reduced artefacts and background noise is obtained. In this study, quantitative evaluation with simulation data and actual clinical data were conducted. As a result, characteristic peaks of evoked potential could be gained removing background EEG and artefacts using suggested shrinkage function. It was improved more than 0.2–1.6Db compared to the conventional averaging method. Also, the system for measuring and analyzing evoked potentials using DSP is implemented.     

An underwater acoustic MAC protocol using reverse opportunistic packet appending

Underwater communication primarily utilizes propagation of acoustic waves in water. Its unique characteristics, including slow propagation speed and low data rates, pose many challenges to Media Access Control (MAC) protocol design. In most existing handshaking-based underwater MAC protocols, only an initiating sender can transmit data packets to its intended receiver after a channel reservation through a Request-to-Send (RTS)/Clear-to-Send (CTS) handshake. This conventional single-node transmission approach is particularly inefficient in underwater environments, as it does not account for long propagation delays. To improve channel utilization in high latency environments, we propose a novel approach that exploits the idle waiting time during a 2-way handshake to set up concurrent transmissions from multiple nodes. The sender can coordinate multiple first-hop neighbors (appenders) to use the current handshake opportunity to transmit (append) their data packets with partially overlapping transmission times. After the sender finishes transmitting its packets to its own receiver, it starts to receive incoming appended packets that arrive in a collision-free packet train. This not only reduces the amount of time spent on control signaling, but it also greatly improves packet exchange efficiency. Based on this idea, we propose an asynchronous, single-channel handshaking-based MAC protocol based on reverse opportunistic packet appending (ROPA). From extensive simulations (single- and multi-hop networks) and comparisons with several existing MAC protocols, including MACA-U, MACA-UPT, BiC-MAC, Slotted-FAMA, DACAP, unslotted Aloha, we show that ROPA significantly increases channel utilization and offers performance gains in throughput and delay while attaining a stable saturation throughput.     

Automated classification of touching or overlapping M-FISH chromosomes by region fusion and homolog pairing

Automated separation and classification of touching or overlapping chromosomes in a metaphase image is a critical step in computer-aided chromosome analysis. The advent of the multiplex fluorescence in situ hybridization (M-FISH) technology enables multi-spectral chromosome image with rich spectral information and DAPI image with abundant texture information. This paper presents a fusion classification scheme to improve the segmentation of overlapping and touching chromosomes. First, the texture and spectral information is fused to partition the chromosome cluster into a series of homologous regions. Then a graph-theoretical classification and pairing method is proposed to resolve any remaining ambiguity of the aforementioned separation process. Experiment results demonstrate that the proposed method outperforms conventional multi-spectral classification methods in touching and overlapping chromosome separation.     

“Area Without Numbers”: Using Touchscreen Dynamic Geometry to Reason About Shape

Abstract

In this article, we report on two lessons aimed at introducing junior high school students to the idea of shearing in a touchscreen dynamic geometry environment. By using shearing, we hoped to shift students’ attention away from a formula-driven, computational conception of area toward a more geometric one. We found that the students were able to solve several problems involving the comparison of polygons based on verbal and diagrammatic explanations. We describe the features of the touchscreen dynamic geometry technology that supported their learning, as well as the specific role that the teacher played in modeling students’ reasoning about area.     

Effects of wheelchair-based rehabilitation on the physical functions and health perception of stroke patients

The purpose of this study was to investigate the effect of wheelchair-based rehabilitation on the physical functions, health perception, and blood lipids according to the length of time since the stroke. Wheelchair-based rehabilitation progressed for 60 min per session, five times per week for 6 weeks. Physical functions, health perception (SF-36), and blood lipids were measured before and after rehabilitation. Physical fitness for physical function such as agility, grip strength, lower-body muscular endurance, flexibility, and posture control significantly improved after 6 weeks of rehabilitation. Health perception improved significantly with physical and mental health according to time since the stroke occurred. Overall, 6 weeks of wheelchair-based rehabilitation had a positive effect on the physical function and health perception regardless of stroke duration. Wheelchair-based rehabilitation had a more positive effect on the physical functions and health perception on the patients who suffered their stroke most recently. We suggested that wheelchair-based rehabilitation is necessary even through wheelchair based for chronic stroke patients, and it was useful their improved of quality of life.     

A scalable algorithm for mining maximal frequent sequences using a sample

In this paper, we propose an efficient scalable algorithm for mining Maximal Sequential Patterns using Sampling (MSPS). The MSPS algorithm reduces much more search space than other algorithms because both the subsequence infrequency-based pruning and the supersequence frequency-based pruning are applied. In MSPS, a sampling technique is used to identify long frequent sequences earlier, instead of enumerating all their subsequences. We propose how to adjust the user-specified minimum support level for mining a sample of the database to achieve better overall performance. This method makes sampling more efficient when the minimum support is small. A signature-based method and a hash-based method are developed for the subsequence infrequency-based pruning when the seed set of frequent sequences for the candidate generation is too big to be loaded into memory. A prefix tree structure is developed to count the candidate sequences of different sizes during the database scanning, and it also facilitates the customer sequence trimming. Our experiments showed MSPS has very good performance and better scalability than other algorithms. Congnan Luo received the B.E. degree in Computer Science from Tsinghua University, Beijing, P.R. China, in 1997, the M.S. degree in Computer Science from the Institute of Software, Chinese Academy of Sciences, Beijing, P.R. China, in 2000, and the Ph.D. degree in Computer Science and Engineering from Wright State University, Dayton, OH, in 2006. Currently he is a technical staff at the Teradata division of NCR in San Diego, CA, and his research interests include data mining, machine learning, and databases. Soon M. Chung received the B.S. degree in Electronic Engineering from Seoul National University, Korea, in 1979, the M.S. degree in Electrical Engineering from Korea Advanced Institute of Science and Technology, Korea, in 1981, and the Ph.D. degree in Computer Engineering from Syracuse University, Syracuse, New York, in 1990. He is currently a Professor in the Department of Computer Science and Engineering at Wright State University, Dayton, OH. His research interests include database, data mining, Grid computing, text mining, XML, and parallel and distributed processing.     

A survey on algorithms for mining frequent itemsets over data streams

The increasing prominence of data streams arising in a wide range of advanced applications such as fraud detection and trend learning has led to the study of online mining of frequent itemsets (FIs). Unlike mining static databases, mining data streams poses many new challenges. In addition to the one-scan nature, the unbounded memory requirement and the high data arrival rate of data streams, the combinatorial explosion of itemsets exacerbates the mining task. The high complexity of the FI mining problem hinders the application of the stream mining techniques. We recognize that a critical review of existing techniques is needed in order to design and develop efficient mining algorithms and data structures that are able to match the processing rate of the mining with the high arrival rate of data streams. Within a unifying set of notations and terminologies, we describe in this paper the efforts and main techniques for mining data streams and present a comprehensive survey of a number of the state-of-the-art algorithms on mining frequent itemsets over data streams. We classify the stream-mining techniques into two categories based on the window model that they adopt in order to provide insights into how and why the techniques are useful. Then, we further analyze the algorithms according to whether they are exact or approximate and, for approximate approaches, whether they are false-positive or false-negative. We also discuss various interesting issues, including the merits and limitations in existing research and substantive areas for future research.