Akara SOPHARAK Bunyarit UYYANONVARA Sarah BARMAN Thomas WILLIAMSON
To prevent blindness from diabetic retinopathy, periodic screening and early diagnosis are neccessary. Due to lack of expert ophthalmologists in rural area, automated early exudate (one of visible sign of diabetic retinopathy) detection could help to reduce the number of blindness in diabetic patients. Traditional automatic exudate detection methods are based on specific parameter configuration, while the machine learning approaches which seems more flexible may be computationally high cost. A comparative analysis of traditional and machine learning of exudates detection, namely, mathematical morphology, fuzzy c-means clustering, naive Bayesian classifier, Support Vector Machine and Nearest Neighbor classifier are presented. Detected exudates are validated with expert ophthalmologists' hand-drawn ground-truths. The sensitivity, specificity, precision, accuracy and time complexity of each method are also compared.
Yi WAN Takuya ASAKA Tatsuro TAKAHASHI
User Generated Content (UGC) VoD services such as YouTube are becoming more and more popular, and their maintenance costs are growing as well. Many P2P solutions have been proposed to reduce server load in such systems, but almost all of them focus on the single-video approach, which only has limited effect on the systems serving short videos such as UGC. The purpose of this paper is to investigate the potential of an alternative approach, the multi-video approach, and we use a very simple method called collaborative caching to show that methods using the multi-video approach are generally more suitable for current UGC VoD systems. We also study the influence of the major design factors through simulations and provide guidelines for efficiently building systems with this method.
Despite the prevalence of Java workloads across a variety of processor architectures, there is very little published data on the impact of the various processor design decisions on Java performance. We attribute the lack of data to the large design space resulting from the complexity of the modern superscalar processor and the additional complexities associated with executing Java bytecode using a virtual machine. To address this shortcoming, we use a statistically rigorous methodology to systematically quantify the the impact of the various processor microarchitecture parameters on Java execution performance. The adopted methodology enables efficient screening of significant factor effects in a large design space consisting of 35 factors (32-billion potential configurations) using merely 72 observations per benchmark application. We quantify and tabulate the significance of each of the 35 factors for 13 benchmark applications. While these tables provide various insights into Java performance, they consistently highlight the performance significance of the instruction delivery mechanism, especially the instruction cache and the ITLB design parameters. Furthermore, these tables enable the architect to identify processor bottlenecks for Java workloads by providing an estimate of the relative impact of various design decisions.
A new random access channel (RACH) preamble detection scheme using variable coherent correlation intervals (CCIs) is proposed. It is shown first that it is enough to employ two CCIs for supporting a user equipment (UE) velocity of 300 km/h, and then a CCI selection criterion is proposed. Computer simulation results indicate that the proposed scheme can provide a robust detection performance in time-varying fading channel environments.
Hirofumi YAMAMOTO Hideo OKUMA Eiichiro SUMITA
In the current statistical machine translation (SMT), erroneous word reordering is one of the most serious problems. To resolve this problem, many word-reordering constraint techniques have been proposed. Inversion transduction grammar (ITG) is one of these constraints. In ITG constraints, target-side word order is obtained by rotating nodes of the source-side binary tree. In these node rotations, the source binary tree instance is not considered. Therefore, stronger constraints for word reordering can be obtained by imposing further constraints derived from the source tree on the ITG constraints. For example, for the source word sequence { a b c d }, ITG constraints allow a total of twenty-two target word orderings. However, when the source binary tree instance ((a b) (c d)) is given, our proposed "imposing source tree on ITG" (IST-ITG) constraints allow only eight word orderings. The reduction in the number of word-order permutations by our proposed stronger constraints efficiently suppresses erroneous word orderings. In our experiments with IST-ITG using the NIST MT08 English-to-Chinese translation track's data, the proposed method resulted in a 1.8-points improvement in character BLEU-4 (35.2 to 37.0) and a 6.2% lower CER (74.1 to 67.9%) compared with our baseline condition.
Huifang FENG Yantai SHU Maode MA
The predictability of network traffic is an important and widely studied topic because it can lead to the solutions to get more efficient dynamic bandwidth allocation, admission control, congestion control and better performance wireless networks. Support vector machine (SVM) is a novel type of learning machine based on statistical learning theory, can solve small-sample learning problems. The work presented in this paper aims to examine the feasibility of applying SVM to predict actual WLAN traffic. We study one-step-ahead prediction and multi-step-ahead prediction without any assumption on the statistical property of actual WLAN traffic. We also evaluate the performance of different prediction models such as ARIMA, FARIMA, artificial neural network, and wavelet-based model using three actual WLAN traffic. The results show that the SVM-based model for predicting WLAN traffic is reasonable and feasible and has the best performance among the above mentioned prediction models.
Mitsuru KAKIMOTO Hisaaki HATANO Yosoko NISHIZAWA
In this paper, we present a forecasting method for the view of Mt. Fuji as an application of Earth observation data (EOD) obtained by satellites. We defined the Mt. Fuji viewing index (FVI) that characterises how well the mountain looks on a given day, based on photo data produced by a fixed-point observation. A long-term predictor of FVI, ranging from 0 to 30 days, was constructed through support vector machine regression on climate and earth observation data. It was found that the aerosol mass concentration (AMC) improves prediction performance, and such performance is particularly significant in the long-term range.
The IEEE 802.16e standard adopts a power-saving mode (PSM) with a truncated binary exponent (TBE) algorithm for determining sleep intervals. Although the TBE algorithm allows more flexibility in determining sleep intervals, it does not consider the network delay of a response packet. In this letter, we suggest PSM that is based on the conditional probability density function (c-pdf) for the response packet's arrival time at the base station (BS). The proposed algorithm determines sleep interval placement so that the response packet may arrive at the BS during each sleep interval with the same conditional probability. The results show that the proposed algorithm outperforms the TBE algorithm with respect to the packet-buffering delay in the BS and the energy consumption of a mobile station (MS).
Jiancheng SUN Chongxun ZHENG Xiaohe LI
With a Gaussian kernel function, we find that the distance between two classes (DBTC) can be used as a class separability criterion in feature space since the between-class separation and the within-class data distribution are taken into account impliedly. To test the validity of DBTC, we develop a method of tuning the kernel parameters in support vector machine (SVM) algorithm by maximizing the DBTC in feature space. Experimental results on the real-world data show that the proposed method consistently outperforms corresponding hyperparameters tuning methods.
Tetsuya KAWANISHI Takahide SAKAMOTO Akito CHIBA
We present recent progress of high-speed Mach-Zehnder modulator technologies for advanced modulation formats. Multi-level quadrature amplitude modulation signal can be synthesized by using parallel Mach-Zehnder modulators. We can generate complicated multi-level optical signals from binary data streams, where binary modulated signals are vectorially summed in optical circuits. Frequency response of each Mach-Zehnder interferometer is also very important to achieve high-speed signals. We can enhance the bandwidth of the response, with thin substrate. 87 Gbaud modulation was demonstrated with a dual-parallel Mach-Zehnder modulator.
Hisashi KASHIMA Tsuyoshi IDE Tsuyoshi KATO Masashi SUGIYAMA
Kernel methods such as the support vector machine are one of the most successful algorithms in modern machine learning. Their advantage is that linear algorithms are extended to non-linear scenarios in a straightforward way by the use of the kernel trick. However, naive use of kernel methods is computationally expensive since the computational complexity typically scales cubically with respect to the number of training samples. In this article, we review recent advances in the kernel methods, with emphasis on scalability for massive problems.
In this letter, we consider a problem of global stabilization of a class of approximately feedback linearized systems. We propose a new nonlinear control approach which includes a nonlinear controller and a Lyapunov-based design method. Our new nonlinear control approach broadens the class of systems under consideration over the existing results.
The electromagnetic fields emitted from an electrostatic discharge (ESD) event occurring between charged metals cause seriously damage high-tech equipment. In order to clarify the generation mechanism of such ESD fields and also to reduce them, we previously proposed a finite-difference time-domain (FDTD) algorithm based on a delta-gap feeding method and a frequency dispersion characteristic formula (Naito's formula) of ferrite material for simulating the ESD fields due to a spark between the charged metals with ferrite core attachment. In the present study, by integrating the above FDTD algorithm and a spark-resistance formula, we simulated both of the ESD itself and the resultant fields for the metal bars with ferrite core attachment, and demonstrated that the core attachment close to the spark gap suppresses the magnetic field level. This finding was also validated via 6-GHz wide-band measurement of the magnetic near-field.
Nobuaki TOJO Nozomu TOGAWA Masao YANAGISAWA Tatsuo OHTSUKI
In an embedded system where a single application or a class of applications is repeatedly executed on a processor, its cache configuration can be customized such that an optimal one is achieved. We can have an optimal cache configuration which minimizes overall memory access time by varying the three cache parameters: the number of sets, a line size, and an associativity. In this paper, we first propose two cache simulation algorithms: CRCB1 and CRCB2, based on Cache Inclusion Property. They realize exact cache simulation but decrease the number of cache hit/miss judgments dramatically. We further propose three more cache design space exploration algorithms: CRMF1, CRMF2, and CRMF3, based on our experimental observations. They can find an almost optimal cache configuration from the viewpoint of access time. By using our approach, the number of cache hit/miss judgments required for optimizing cache configurations is reduced to 1/10-1/50 compared to conventional approaches. As a result, our proposed approach totally runs an average of 3.2 times faster and a maximum of 5.3 times faster compared to the fastest approach proposed so far. Our proposed cache simulation approach achieves the world fastest cache design space exploration when optimizing total memory access time.
This letter proposes a simple modification of LEACH protocol to exploit its multi-hop scenario for user cooperation. Instead of a single cluster-head we propose M cluster-heads in each cluster to obtain the diversity of order M. All cluster-heads gather data from all sensor nodes within the cluster using the same technique as LEACH. Cluster-heads transmit gathered data cooperatively towards the destination or higher order cluster-head. We propose a code combining based cooperative protocol. We also develop the upper bounds on frame error rate (FER) for our proposal. Simulation and analysis show that our proposal can significantly prolong the system lifetime.
Yoshihisa KISHIYAMA Kenichi HIGUCHI Mamoru SAWAHASHI
This paper presents the optimum physical random access channel (PRACH) structure in terms of the number of control signaling bits accommodated and the transmission bandwidth based on the link budget in order to satisfy the coverage requirement for the single-carrier (SC)-FDMA based E-UTRA uplink. First, we present the design concept of the PRACH structure considering the purposes of the random access procedure in the E-UTRA. Simulation evaluations including a system-level simulation show that a PRACH comprising a 0.5-msec preamble sequence can convey a 6-bit control signal at the cell edge when the inter-site distance (ISD) is 500 m under full channel load conditions with one-cell frequency reuse. It is also shown, however, that a PRACH longer than one-sub-frame, e.g., 1.0 msec, is necessary to support the ISD of 1732 m assuming the same conditions. We also show that the best transmission bandwidth for the PRACH is approximately 1.08-4.5 MHz from the viewpoint of the misdetection probability, and a 1.08-MHz transmission bandwidth is suitable considering other aspects such as flexible resource assignment in the time domain and a small number of options in the transmission bandwidth.
Tomoko IZUMI Taisuke IZUMI Fukuhito OOSHITA Hirotsugu KAKUGAWA Toshimitsu MASUZAWA
Biologically-inspired approaches are one of the most promising approaches to realize highly-adaptive distributed systems. Biological systems inherently have self-* properties, such as self-stabilization, self-adaptation, self-configuration, self-optimization and self-healing. Thus, the application of biological systems into distributed systems has attracted a lot of attention recently. In this paper, we present one successful result of bio-inspired approach: we propose distributed algorithms for resource replication inspired by the single species population model. Resource replication is a crucial technique for improving system performance of distributed applications with shared resources. In systems using resource replication, generally, a larger number of replicas lead to shorter time to reach a replica of a requested resource but consume more storage of the hosts. Therefore, it is indispensable to adjust the number of replicas appropriately for the resource sharing application. This paper considers the problem for controlling the densities of replicas adaptively in dynamic networks and proposes two bio-inspired distributed algorithms for the problem. In the first algorithm, we try to control the replica density for a single resource. However, in a system where multiple resources coexist, the algorithm needs high network cost and the exact knowledge at each node about all resources in the network. In the second algorithm, the densities of all resources are controlled by the single algorithm without high network cost and the exact knowledge about all resources. This paper shows by simulations that these two algorithms realize self-adaptation of the replica density in dynamic networks.
Gi-Ho PARK Jung-Wook PARK Hoi-Jin LEE Gunok JUNG Sung-Bae PARK Shin-Dug KIM
This paper presents a cache way enabling mechanism using branch target addresses. This mechanism uses branch prediction information to avoid the power consumption due to unnecessary cache way access by enabling only the cache way(s) that should be accessed. The proposed cache way enabling mechanism reduces the power consumption of the instruction cache by 63% without any performance degradation of the processor. An ARM1136 processor simulator and the Synopsys PrimeTime are used to perform the performance/power simulation and static timing analysis of the proposed mechanisms respectively.
Rameswar DEBNATH Masakazu MURAMATSU Haruhisa TAKAHASHI
The core of the support vector machine (SVM) problem is a quadratic programming problem with a linear constraint and bounded variables. This problem can be transformed into the second order cone programming (SOCP) problems. An interior-point-method (IPM) can be designed for the SOCP problems in terms of storage requirements as well as computational complexity if the kernel matrix has low-rank. If the kernel matrix is not a low-rank matrix, it can be approximated by a low-rank positive semi-definite matrix, which in turn will be fed into the optimizer. In this paper we present two SOCP formulations for each SVM classification and regression problem. There are several search direction methods for implementing SOCPs. Our main goal is to find a better search direction for implementing the SOCP formulations of the SVM problems. Two popular search direction methods: HKM and AHO are tested analytically for the SVM problems, and efficiently implemented. The computational costs of each iteration of the HKM and AHO search direction methods are shown to be the same for the SVM problems. Thus, the training time depends on the number of IPM iterations. Our experimental results show that the HKM method converges faster than the AHO method. We also compare our results with the method proposed in Fine and Scheinberg (2001) that also exploits the low-rank of the kernel matrix, the state-of-the-art SVM optimization softwares SVMTorch and SVMlight. The proposed methods are also compared with Joachims 'Linear SVM' method on linear kernel.
In this article, we introduce a new concept for the future information environment, called an "ambient information environment (AmIE)." We first explain it, especially emphasizing the difference from the existing ubiquitous information environment (UbIE), which is an interaction between users and environments. Then, we focus on an ambient networking environment (AmNE) which supports the AmIE as a networking infrastructure. Our approach of a biologically inspired framework is next described in order to demonstrate why such an approach is necessary in the AmIE. Finally, we show some example for building the AmNE.