Tree structured data often appear in bioinformatics. For example, glycans, RNA secondary structures and phylogenetic trees usually have tree structures. Comparison of trees is one of fundamental tasks in analysis of these data. Various distance measures have been proposed and utilized for comparison of trees, among which extensive studies have been done on tree edit distance. In this paper, we review key results and our recent results on the tree edit distance problem and related problems. In particular, we review polynomial time exact algorithms and more efficient approximation algorithms for the edit distance problem for ordered trees, and approximation algorithms for the largest common sub-tree problem for unordered trees. We also review applications of tree edit distance and its variants to bioinformatics with focusing on comparison of glycan structures.

抄録全体を表示

A framework of context-sensitive grammar transform for speeding-up compressed pattern matching (CPM) is proposed. A greedy compression algorithm with the transform model is presented as well as a Knuth-Morris-Pratt (KMP)-type compressed pattern matching algorithm. The compression ratio is a match for gzip and Re-Pair, and the search speed of our CPM algorithm is almost twice faster than the KMP-type CPM algorithm on Byte-Pair-Encoding by Shibata et al.[18], and in the case of short patterns, faster than the Boyer-Moore-Horspool algorithm with the stopper encoding by Rautio et al.[14], which is regarded as one of the best combinations that allows a practically fast search.

抄録全体を表示

In this paper, we obtain some refinement of representation theorems for context-free languages by using Dyck languages, insertion systems, strictly locally testable languages, and morphisms. For instance, we improved the Chomsky-Schützenberger representation theorem and show that each context-free language L can be represented in the form L = h (D ∩ R), where D is a Dyck language, R is a strictly 3-testable language, and h is a morphism. A similar representation for context-free languages can be obtained, using insertion systems of weight (3, 0) and strictly 4-testable languages.

抄録全体を表示

In this paper, we study lower bounds on the query complexity of testing algorithms for various problems. Given an oracle that returns information of an input object, a testing algorithm distinguishes the case that the object has a given property P from the case that it has a large distance to having P with probability at least 2/3. The query complexity of an algorithm is measured by the number of accesses to the oracle. We introduce two reductions that preserve the query complexity. One is derived from the gap-preserving local reduction and the other is from the L-reduction. By using the former reduction, we show linear lower bounds on the query complexity for testing basic NP-complete properties, i.e., 3-edge-colorability, directed Hamiltonian path/cycle, undirected Hamiltonian path/cycle, 3-dimensional matching and NP-complete generalized satisfiability problems. Also, using the second reduction, we show a linear lower bound on the query complexity of approximating the size of the maximum 3-dimensional matching.

抄録全体を表示

This paper deals with the maximum-weight 2-path packing problem (M2PP), which is the problem of computing a set of vertex-disjoint paths of length 2 in a given edge-weighted complete graph so that the total weight of edges in the paths is maximized. Previously, Hassin and Rubinstein gave a randomized cubic-time approximation algorithm for M2PP which achieves an expected ratio of 35/67 - ε ≈ 0.5223 - ε for any constant ε > 0. We refine their algorithm and derandomize it to obtain a deterministic cubic-time approximation algorithm for the problem which achieves a better ratio (namely, 0.5265 - ε for any constant ε > 0).

抄録全体を表示

In this paper, we address the following problems: Given a sequence A of n real numbers, and four parameters I, J, X and Y with I ≤ J and X ≤ Y, find the longest (or shortest) subsequence of A such that its length is between I and J and its sum is between X and Y. We present an online and an offline algorithm for the problems, both run in O(n log n) time, which are optimal.

抄録全体を表示

An inverter is a circuit which outputs ¬x₁, ¬x₂, …, ¬x_n for any Boolean inputs x₁, x₂, …, x_n. We consider constructing an inverter with AND gates and OR gates and a few NOT gates. Beals, Nishino and Tanaka have given a construction of an inverter which has size O(nlog n) and depth O(log n) and uses ⌈ log(n+1) ⌉ NOT gates. In this paper we give a construction of an inverter which has size O(n) and depth log^1+o(1)n and uses log^1+o(1)n NOT gates. This is the first negation-limited inverter of linear size using only o(n) NOT gates. We also discuss implications of our construction for negation-limited circuit complexity.

抄録全体を表示

A (k, δ, ε)-locally decodable code $C:{\bf F}_{q}^{n} \ ightarrow {\bf F}_{q}^{N}$ is an error-correcting code that encodes $\vec{x}=(x_{1},x_{2},\ldots,x_{n}) \in {\bf F}_{q}^{n}$ to $C(\vec{x}) \in {\bf F}_{q}^{N}$ and has the following property: For any $\vec{y} \in {\bf F}_{q}^{N}$ such that $d(\vec{y},C(\vec{x})) \leq \delta N$ and each 1 ≤ i ≤ n, the symbol x_i of $ of $\vec{x}$ can be recovered with probability at least 1 - ε by a randomized decoding algorithm looking at only k coordinates of $\vec{y}$. The efficiency of a (k, δ, ε)-locally decodable code $C:{\bf F}_{q}^{n} \ ightarrow {\bf F}_{q}^{N}$ is measured by the code length N and the number k of queries. For a k-query locally decodable code $C:{\bf F}_{q}^{n} \ ightarrow {\bf F}_{q}^{N}$, the code length N was conjectured to be exponential of n, i.e., N = exp(n^Ω(1)), however, this was disproved. Yekhanin [In Proc. of STOC, 2007] showed that there exists a 3-query locally decodable code $C:{\bf F}_{2}^{n} \ ightarrow {\bf F}_{2}^{N}$ such that N=exp(n^{1/log log n}) assuming that infinitely many Mersenne primes exist. For a 3-query locally decodable code $C:{\bf F}_{q}^{n} \ ightarrow {\bf F}_{q}^{N}$, Efremenko [ECCC Report No.69, 2008] further reduced the code length to $N=\exp(n^{O((\log \log n/ \log n)^{1/2})})$, and in general showed that for any integer r > 1, there exists a 2^r-query locally decodable code $C:{\bf F}_{q}^{n} \ ightarrow {\bf F}_{q}^{N}$ such that $N=\exp(n^{O((\log \log n/ \log n)^{1-1/r})})$. In this paper, we will present improved constructions for query-efficient locally decodable codes by introducing a technique of “composition of locally decodable codes,” and show that for any integer r > 5, there exists a 9 · 2^r-4-query locally decodable code $C:{\bf F}_{q}^{n} \ ightarrow {\bf F}_{q}^{N}$ such that $N=\exp(n^{O((\log \log n/ \log n)^{1-1/r})})$.

抄録全体を表示

In [1] and [2] we have presented a simulation-based algorithm for optimizing the average reward in a parameterized continuous-time, finite-state semi-Markov Decision Process (SMDP). We approximated the gradient of the average reward. Then, a simulation-based algorithm was proposed to estimate the approximate gradient of the average reward (called GSMDP), using only a single sample path of the underlying Markov chain. GSMDP was proved to converge with probability 1. In this paper, we give bounds on the approximation and estimation errors for GSMDP algorithm. The approximation error of that approximation is the size of the difference between the true gradient and the approximate gradient. The estimation error, the size of the difference between the output of the algorithm and its asymptotic output, arises because the algorithm sees only a finite data sequence.

抄録全体を表示

We introduce a complexity measure r for the class $\mathcal{F}$ of read-once formulas over the basis {AND, OR, NOT, XOR, MUX} and show that for any Boolean formula F in the class $\mathcal{F}$, r(F) is a lower bound on the quantum query complexity of the Boolean function that F represents. We also show that for any Boolean function f represented by a formula in $\mathcal{F}$, the deterministic query complexity of f is only quadratically larger than the quantum query complexity of f. Thus, the paper gives further evidence for the conjecture that there is an only quadratic gap for all functions.

抄録全体を表示

For a graph G, a biclique edge partition S_BP(G) is a collection of bicliques (complete bipartite subgraphs) B_i such that each edge of G is contained in exactly one B_i. The Minimum Biclique Edge Partition Problem (MBEPP) asks for S_BP(G) with the minimum size. In this paper, we show that for arbitrary small ε > 0, (6053/6052 - ε)-approximation of MBEPP is NP-hard.

抄録全体を表示

Structured light vision systems are based on the fact that the pixel location of light in an image has a unique association with the object depth. However, their applications are mainly limited to the properties of the object surface and the lighting conditions. This paper presents a robust vision system for accurate acquisition of 3-D surface data based on optimal structured light. To achieve depth measurement for a dynamic scene, the data acquisition must be performed with only a single image. Our special arrangement makes the image of the light stripe remaining sharp while the background becomes blurred. Moreover, a dynamic programming approach is proposed to track the optimal path while the laser beam is invisible or divergent under extreme condition. The principle and necessary mathematics for implementing the algorithm are presented. The robustness of the system against uncalibrated errors is demonstrated.

抄録全体を表示

In this paper, a fast and memory-efficient VLSI architecture for output probability computations of continuous Hidden Markov Models (HMMs) is presented. These computations are the most time-consuming part of HMM-based recognition systems. High-speed VLSI architectures with small registers and low-power dissipation are required for the development of mobile embedded systems with capable human interfaces. We demonstrate store-based block parallel processing (StoreBPP) for output probability computations and present a VLSI architecture that supports it. When the number of HMM states is adequate for accurate recognition, compared with conventional stream-based block parallel processing (StreamBPP) architectures, the proposed architecture requires fewer registers and processing elements and less processing time. The processing elements used in the StreamBPP architecture are identical to those used in the StoreBPP architecture. From a VLSI architectural viewpoint, a comparison shows the efficiency of the proposed architecture through efficient use of registers for storing input feature vectors and intermediate results during computation.

抄録全体を表示

This paper presents a proposal of an expected-credibility-based job scheduling method for volunteer computing (VC) systems with malicious participants who return erroneous results. Credibility-based voting is a promising approach to guaranteeing the computational correctness of VC systems. However, it relies on a simple round-robin job scheduling method that does not consider the jobs' order of execution, thereby resulting in numerous unnecessary job allocations and performance degradation of VC systems. To improve the performance of VC systems, the proposed job scheduling method selects a job to be executed prior to others dynamically based on two novel metrics: expected credibility and the expected number of results for each job. Simulation of VCs shows that the proposed method can improve the VC system performance up to 11%; It always outperforms the original round-robin method irrespective of the value of unknown parameters such as population and behavior of saboteurs.

抄録全体を表示

Privacy preserving association rule mining algorithms have been designed for discovering the relations between variables in data while maintaining the data privacy. In this article we revise one of the recently introduced schemes for association rule mining using fake transactions (FS). In particular, our analysis shows that the FS scheme has exhaustive storage and high computation requirements for guaranteeing a reasonable level of privacy. We introduce a realistic definition of privacy that benefits from the average case privacy and motivates the study of a weakness in the structure of FS by fake transactions filtering. In order to overcome this problem, we improve the FS scheme by presenting a hybrid scheme that considers both privacy and resources as two concurrent guidelines. Analytical and empirical results show the efficiency and applicability of our proposed scheme.

抄録全体を表示

This paper presents a novel approach for robot localization using landmark maps. With recent progress in SLAM researches, it has become crucial for a robot to obtain and use large-size maps that are incrementally built by other mapper robots. Our localization approach successfully works with such incremental and large-size maps. In literature, RANSAC map-matching has been a promising approach for large-size maps. We extend the RANSAC map-matching so as to deal with incremental maps. We combine the incremental RANSAC with an incremental LSH database and develop a hybrid of the position-based and the appearance-based approaches. A series of experiments using radish dataset show promising results.

抄録全体を表示

This paper proposes a topic-independent method for automatically scoring essay content. Unlike conventional topic-dependent methods, it predicts the human-assigned score of a given essay without training essays written to the same topic as the target essay. To achieve this, this paper introduces a new measure called MIDF that measures how important and relevant a word is in a given essay. The proposed method predicts the score relying on the distribution of MIDF. Surprisingly, experiments show that the proposed method achieves an accuracy of 0.848 and performs as well as or even better than conventional topic-dependent methods.

抄録全体を表示

Traditional wavelet-based speech enhancement algorithms are ineffective in the presence of highly non-stationary noise because of the difficulties in the accurate estimation of the local noise spectrum. In this paper, a simple method of noise estimation employing the use of a voice activity detector is proposed. We can improve the output of a wavelet-based speech enhancement algorithm in the presence of random noise bursts according to the results of VAD decision. The noisy speech is first preprocessed using bark-scale wavelet packet decomposition (BSWPD) to convert a noisy signal into wavelet coefficients (WCs). It is found that the VAD using bark-scale spectral entropy, called as BS-Entropy, parameter is superior to other energy-based approach especially in variable noise-level. The wavelet coefficient threshold (WCT) of each subband is then temporally adjusted according to the result of VAD approach. In a speech-dominated frame, the speech is categorized into either a voiced frame or an unvoiced frame. A voiced frame possesses a strong tone-like spectrum in lower subbands, so that the WCs of lower-band must be reserved. On the contrary, the WCT tends to increase in lower-band if the speech is categorized as unvoiced. In a noise-dominated frame, the background noise can be almost completely removed by increasing the WCT. The objective and subjective experimental results are then used to evaluate the proposed system. The experiments show that this algorithm is valid on various noise conditions, especially for color noise and non-stationary noise conditions.

抄録全体を表示

The precise noise modeling of complementary metal oxide semiconductor image sensor (CMOS image sensor: CIS) is a significant key in understanding the noise source mechanisms, optimizing sensor design, designing noise reduction circuit, and enhancing image quality. Therefore, this paper presents an accurate random telegraph signal (RTS) noise analysis model and a novel quantitative evaluation method in motion picture for the visual sensory evaluation of CIS. In this paper, two main works will be introduced. One is that the exposure process of a video camera is simulated, in which a Gaussian noise and an RTS noise in pinned-photodiode CMOS pixels are modeled in time domain and spatial domain; the other is that a new video quality evaluation method for RTS noise is proposed. Simulation results obtained reveal that the proposed noise modeling for CIS can approximate its physical process and the proposed video quality evaluation method for RTS noise performs effectively as compared to other evaluation methods. Based on the experimental results, conclusions on how the spatial distribution of an RTS noise affects the quality of motion picture are carried out.

抄録全体を表示

In this paper, Cellular Neural Networks using genetic algorithm (GA-CNNs) are designed for CMOS image noise reduction. Cellular Neural Networks (CNNs) could be an efficient way to apply to the image processing technique, since CNNs have high-speed parallel signal processing characteristics. Adaptive CNNs structure is designed for the reduction of Photon Shot Noise (PSN) changed according to the average number of photons, and the design of templates for adaptive CNNs is based on the genetic algorithm using real numbers. These templates are optimized to suppress PSN in corrupted images. The simulation results show that the adaptive GA-CNNs more efficiently reduce PSN than do the other noise reduction methods and can be used as a high-quality and low-cost noise reduction filter for PSN. The proposed method is designed for real-time implementation. Therefore, it can be used as a noise reduction filter for many commercial applications. The simulation results also show the feasibility to design the CNNs template for a variety of problems based on the statistical image model.

抄録全体を表示

A band-pass bilateral filter is an improved variant of a bilateral filter that does not have low-pass characteristics but has band-pass characteristics. Unfortunately, its computation time is relatively large since all pixels are subjected to Gaussian calculation. To solve this problem, we pay attention to a nonlinear filter called ε-filter and propose an advanced ε-filter labeled band-pass ε-filter. As ε-filter has low-pass characteristics due to spatial filtering, it does not enhance the image contrast. On the other hand, band-pass ε-filter does not have low-pass characteristics but has band-pass characteristics to enhance the image contrast around edges unlike ε-filter. The filter works not only as a noise reduction filter but also as an edge detection filter depending on the filter setting. Due to its simple design, the calculation cost is relatively small compared to the band-pass bilateral filter. To show the effectiveness of the proposed method, we report the results of some comparison experiments on the filter characteristics and computational cost.

抄録全体を表示

Fractals are mathematical or natural objects that are made of parts similar to the whole in certain ways. In this paper a software reliability forecasting method of software failure is proposed based on predictability of fractal time series. The empirical failure data (three data sets of Musa's) are used to demonstrate the performance of the reliability prediction. Compared with other methods, our method is effective.

抄録全体を表示

This paper presents a selective scan slice grouping technique for test data compression. In conventional selective encoding methods, the existence of a conflict bit contributes to large encoding data. However, many conflict bits are efficiently removed using the scan slice grouping technique, which leads to a dramatic improvement of encoding efficiency. Experiments performed with large ITC'99 benchmark circuits presents the effectiveness of the proposed technique and the test data volume is reduced up to 92% compared to random-filled test patterns.

抄録全体を表示

To detect the natural clusters for irregularly shaped data distribution is a difficult task in pattern recognition. In this study, we propose an efficient clustering algorithm for irregularly shaped clusters based on the advantages of spectral clustering and Affinity Propagation (AP) algorithm. We give a new similarity measure based on neighborhood dispersion analysis. The proposed algorithm is a simple but effective method. The experimental results on several data sets show that the algorithm can detect the natural clusters of input data sets, and the clustering results agree well with that of human judgment.

抄録全体を表示

A scalable speech codec consisting of a harmonic codec as the core layer and MDCT-based transform codec as the enhancement layer is often required to provide both very low-rate core communication and fine granular scalability. This structure, however, has a serious drawback for practical use because a time delay caused by transform in each layer is accumulated, resulting in a long overall codec delay. In this letter, a new MDCT structure is proposed to reduce the overall codec delay by eliminating the accumulation of time delay by each transform. In the proposed structure, the time delay is first reduced by forcing two transforms to share a common look-ahead. The error components of MDCT caused by the look-ahead sharing are then analyzed and compensated in the decoder, resulting in perfect reconstruction. The proposed structure reduces the codec delay by the frame size, with an equivalent coding efficiency.

抄録全体を表示

The selection of effective features is especially important in achieving highly accurate speech recognition. Although the mel-cepstrum is a popular and effective feature for speech recognition, it is still unclear that the filterbank adopted in the mel-cepstrum always produces the optimal performance regardless of the phonetic environment of any specific speech recognition task. In this paper, we propose a new cepstral domain feature extraction approach utilizing the entropic distance-based filterbank for highly accurate speech recognition. Experimental results showed that the cepstral features employing the proposed filterbank reduce the relative error by 31% for clean as well as noisy speech compared to the mel-cepstral features.

抄録全体を表示

To improve speech recognition performance, acoustic feature transformation based on discriminant analysis has been widely used. For the same purpose, discriminative training of HMMs has also been used. In this letter we investigate the effectiveness of these two techniques and their combination. We also investigate the robustness of matched and mismatched noise conditions between training and evaluation environments.

抄録全体を表示

A secret image transmission scheme based on vector quantization (VQ) and a secret codebook is proposed in this article. The goal of this scheme is to transmit a set of good-quality images secretly via another high-quality cover image with the same image size. In order to reduce the data size of secret images, the images are encoded by an adaptive codebook. To guarantee the visual quality of secret images, the adaptive codebook applied at the transmitter is transmitted to the receiver secretly as well. Moreover, to enhance the security of the proposed scheme and to compact the data size of the codebook, the adaptive codebook is encoded based on VQ using another codebook generated from the cover image. Experiments show impressive results.

抄録全体を表示

In this letter, we learned overcomplete filters to model rich priors of nature images. Our approach extends the Gaussian Scale Mixture Fields of Experts (GSM FOE), which is a fast approximate model based on Fields of Experts (FOE). In these previous image prior model, the overcomplete case is not considered because of the heavy computation. We introduce the assumption of quasi-orthogonality to the GSM FOE, which allows us to learn overcomplete filters of nature images fast and efficiently. Simulations show these obtained overcomplete filters have properties similar with those of Fields of Experts', and denoising experiments also show the superiority of our model.

抄録全体を表示