• 인터넷정보학회논문지
• /
• 제9권6호
• /
• pp.27-35
• /
• 2008

무선 장치의 확산으로, 무선 애드 혹 망(MANETs, Mobile Ad-hoc NETworks)은 매우 흥미롭고 중요한 기술이 되고 있다. 그러나 MANET은 유선망 보다 더 견고하지 못하다. 유선망을 위하여 설계된 기존의 보안 메커니즘은 새로운 패러다임에서 재 설계되어야 한다. 본 논문에서, 우리는 MANET에서 비정상 행위 탐지 문제를 논의한다. 우리의 연구의 초점은 새로운 또는 알려지지 않은 공격을 탐지할 수 있는 비정상 행위 탐지 모델을 자동적으로 구축하는 기법에 있다. 제안하는 방법은 정상 트래픽에서 특징간 상관 관계 패턴을 포착하기 위하여 러프 집합에 기초한 교차 특징 분석을 수행한다. 제안하는 방법의 성능은 시뮬레이션을 통하여 평가되었다. 그 결과, 제안하는 방법의 성능이 특징 속성값의 확률에 기반 하는 교차 특징 분석을 사용하는 Huang의 방법 보다 성능이 우수함을 보였다. 따라서 제안하는 방법이 비정상 행위를 효율적으로 탐지한다는 것을 알 수 있었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권8호
• /
• pp.2101-2123
• /
• 2023

Recent studies have shown that the neural network-based binary code similarity detection technology performs well in vulnerability mining, plagiarism detection, and malicious code analysis. However, existing cross-architecture methods still suffer from insufficient feature characterization and low discrimination accuracy. To address these issues, this paper proposes a cross-architecture binary function similarity detection method based on composite feature model (SDCFM). Firstly, the binary function is converted into vector representation according to the proposed composite feature model, which is composed of instruction statistical features, control flow graph structural features, and application program interface calling behavioral features. Then, the composite features are embedded by the proposed hierarchical embedding network based on a graph neural network. In which, the block-level features and the function-level features are processed separately and finally fused into the embedding. In addition, to make the trained model more accurate and stable, our method utilizes the embeddings of predecessor nodes to modify the node embedding in the iterative updating process of the graph neural network. To assess the effectiveness of composite feature model, we contrast SDCFM with the state of art method on benchmark datasets. The experimental results show that SDCFM has good performance both on the area under the curve in the binary function similarity detection task and the vulnerable candidate function ranking in vulnerability search task.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권2호
• /
• pp.810-831
• /
• 2019

Recently, continuous dimensional emotion recognition from audiovisual clues has attracted increasing attention in both theory and in practice. The large amount of data involved in the recognition processing decreases the efficiency of most bimodal information fusion algorithms. A novel algorithm, namely the incomplete Cholesky decomposition based kernel cross factor analysis (ICDKCFA), is presented and employed for continuous dimensional audiovisual emotion recognition, in this paper. After the ICDKCFA feature transformation, two basic fusion strategies, namely feature-level fusion and decision-level fusion, are explored to combine the transformed visual and audio features for emotion recognition. Finally, extensive experiments are conducted to evaluate the ICDKCFA approach on the AVEC 2016 Multimodal Affect Recognition Sub-Challenge dataset. The experimental results show that the ICDKCFA method has a higher speed than the original kernel cross factor analysis with the comparable performance. Moreover, the ICDKCFA method achieves a better performance than other common information fusion methods, such as the Canonical correlation analysis, kernel canonical correlation analysis and cross-modal factor analysis based fusion methods.

• Smart Structures and Systems
• /
• 제16권1호
• /
• pp.213-222
• /
• 2015

Early detection and precise location of leakage is of great importance for life-cycle maintenance and management of municipal pipeline system. In the past few years, acoustic emission (AE) techniques have demonstrated to be an excellent tool for on-line leakage detection. Regarding the multi-mode and frequency dispersion characteristics of AE signals propagating along a pipeline, the direct cross-correlation technique that assumes the constant AE propagation velocity does not perform well in practice for acoustic leak location. This paper presents an improved cross-correlation method based on wavelet transform, with due consideration of the frequency dispersion characteristics of AE wave and the contribution of different mode. Laboratory experiments conducted to simulate pipeline gas leakage and investigate the frequency spectrum signatures of AE leak signals. By comparing with the other methods for leak location identification, the feasibility and superiority of the proposed method are verified.

• Journal of electromagnetic engineering and science
• /
• 제18권3호
• /
• pp.206-211
• /
• 2018

In this paper, we proposed a two-level feature vector fusion technique to improve the performance of target classification. The proposed method combines feature vectors of the early-time region and late-time region in the first-level fusion. In the second-level fusion, we combine the monostatic and bistatic features obtained in the first level. The radar cross section (RCS) of the 3D full-scale model is obtained using the electromagnetic analysis tool FEKO, and then, the feature vector of the target is extracted from it. The feature vector based on the waveform structure is used as the feature vector of the early-time region, while the resonance frequency extracted using the evolutionary programming-based CLEAN algorithm is used as the feature vector of the late-time region. The study results show that the two-level fusion method is better than the one-level fusion method.

• International journal of advanced smart convergence
• /
• 제11권2호
• /
• pp.38-43
• /
• 2022

In this study, there are many difficulties in defining and judging creative people because there is no systematic analysis method using accurate standards or numerical values. Analyze and judge whether In the previous study, A study on the application of rule success cases through machine learning algorithm extraction, a case study was conducted to help verify or confirm the psychological personality test and aptitude test. We proposed a solution to a research problem in psychology using machine learning algorithms, Data Mining's Cross Industry Standard Process for Data Mining, and CRISP-DM, which were used in previous studies. After that, this study proposes a solution that helps to judge creative people by applying the feature selection algorithm. In this study, the accuracy was found by using seven feature selection algorithms, and by selecting the feature group classified by the feature selection algorithms, and the result of deriving the classification result with the highest feature obtained through the support vector machine algorithm was obtained.

• 대한안전경영과학회지
• /
• 제8권2호
• /
• pp.51-59
• /
• 2006

So far, there is no technique of quantitative evaluation on danger related to industrial accidents. Therefore, as an endeavor for obtaining technique of quantitative evaluation, this study presents feature analysis of industrial accidents in manufacturing field using QUEST algorithm. In order to analyze feature of industrial accidents, a retrospective analysis was performed in 10,536 subjects (10,313 injured people, 223 deaths). The sample for this work chosen from data related to manufacturing businesses during three years $(2002\sim2004)$ in Korea. The analysis results were very informative since those enable us to know the most important variables such as occurrence type, company size, and occurrence time which can affect injured people. Also, it is found that classification using QUEST algorithm which was performed in this study is very reliable.

• Steel and Composite Structures
• /
• 제35권1호
• /
• pp.29-41
• /
• 2020

Cross-beams of modern through truss bridges are connected to truss chord at its nodes and between them. It results in variable rotational end restraint for cross-beams, thus variable bending moment distribution. This feature is captured in three-dimensional modelling of through truss bridge structure. However, for preliminary design or rapid assessment of service load effects such technique of analysis may not be available. So an analytical method of assessment of rotational end restraint for cross-beam of through truss bridges was worked out. Two cases - nodal cross-beam and inter-nodal cross-beam - were analyzed. Flexural and torsional stiffness of truss members, flexural stiffness of deck members and axial stiffness of wind bracing members in the vicinity of the analyzed cross-beam were taken into account. The provision for reduced stiffness of the X-type wind bracing was made. Finally, general formula for assessment of rotational end restraint was given. Rotational end restraints for cross-beams of three railway through truss bridges were assessed basing on the analytical method and the finite element method (three-dimensional beam-element modelling). Results of both methods show good agreement. The analytical method is able to reflect effects of some structural irregularities. On the basis of the obtained results the general values of rotational end restraint for nodal and inter-nodal cross-beams of railway through truss bridges were suggested.

• 비교문화연구
• /
• 제39권
• /
• pp.381-406
• /
• 2015

Russian nasal consonants / m /, / n / have a feature value not only [+consonant] in common with obstruents, but also [+sonorant] in common with vowels. Nasal / m /(bi-labial) and / n /(dental) have the same place of articulation but different manner of articulation. The feature value of / m / is [+cons, +son, +nas, +ant, -cor, -high, -low, -back, -cont, -del, rel, -strid, +voic], and that of / n / is [+cons, +son, +nas, +ant, +cor, -high, -low, -back, -cont, -del, rel, -strid, + voic]. There is a difference in feature [cor] value of / m / and / n /. In this study it is confirmed that it is a fact that the Russian nasal consonants behave differently from the other consonants in each phonological phenomenon due to their phonological characteristics. The preceding voiced obstruent is changed to an unvoiced one in a process where the last voiceless obstruent in the consonant cluster ' voiced obstruent + nasal /m/ + voiceless obstruent' skips the nasal consonant and spreads its feature value to the preceding voiced obstruent transparently because of the feature [+sonorant] of the nasal consonant. The coronal nasal /n/ participates in a palatalization with the following palatal actively and palatalize preceding plain consonants passively because of markedness hierarchy such as 'Velar > Labial > Coronal'. But the labial nasal /m/ is palatalized with the following velar palatal actively and participates in a palatalization with the following coronal palatal passively. This result helps us confirm the phonological difference of /m/ and /n/ in a palatalization. When the a final consonant is nasal, the unvoicing phenomenon of a final consonant doesn't occur. In such a case as cluster 'obstruent + nasal' the feature value [voiced] of the preceding obstruent doesn't change, but the following nasal can assimilate into the preceding obstruent. When continuing the same nasals / -nn- / in a consonant cluster, the feature value [+cont] of a weak position leads the preceding nasal / n / to be changed into [-cont] / l /. Through the analysis of the frequency of occurrences of consonants in syllabic onsets and codas that should observe the 'Sonority Sequence Principle', the sonority hierarchy of nasal consonants has been confirmed. In a diachronic perspective following nasal / m /, / n / there is a loss of the preceding labial stop and dental stop. But in clusters with the velar stop+nasal, the two-component cluster has been kept phonetically intact.

• 한국멀티미디어학회논문지
• /
• 제18권8호
• /
• pp.901-908
• /
• 2015

Methods of detecting, describing, matching image features, like corners and blobs, have been actively studied as a fundamental step for image processing and computer vision applications. As one of feature description/matching methods, LLAH(Locally Likely Arrangement Hashing) describes image features based on the geometric relationship between their neighbors, and thus is suitable for scenes with poor texture. This paper presents a modified LLAH algorithm, which includes the image features themselves for robustly describing the geometric relationship unlike the original LLAH, and employes a voting-based feature matching scheme that makes feature description much simpler. Then, this paper quantitatively analyzes its performance with synthetic images in the presence of Gaussian noise.