• Journal of the Institute of Convergence Signal Processing
• /
• v.14 no.4
• /
• pp.217-221
• /
• 2013

Recently, researches for methods of generating binary feature descriptors have been actively done. Since matching of binary feature descriptors uses Hamming distance which is based on bit operations, it is much more efficient than that of previous general feature descriptors which uses Euclidean distance based on real number operations. However, since increase in the number of features linearly drops matching speed, in applications such as object tracking where real-time applicability is a must, there has been an increasing demand for methods of further improving the matching speed of binary feature descriptors. In this regard, this paper proposes a method that improves the matching speed greatly while maintaining the matching accuracy by splitting high dimensional binary feature descriptors to several low dimensional ones and sequentially analyzing their partial Hamming distances. To evaluate the efficiency of the proposed method, experiments of comparison with previous matching methods are conducted. In addition, this paper discusses schemes of generating binary feature descriptors for maximizing the performance of the proposed method. Based on the analysis on the performance of several generation schemes, we try to find out the most effective scheme.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.8
• /
• 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.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1995.10b
• /
• pp.218-223
• /
• 1995

A new design scheme of a binary decision tree is proposed. In this scheme a binary decision tree is constructed by using genetic algorithm and FCM algorithm. At each node optimal or near-optimal feature or feature subset among all the available features is selected based on fitness function in genetic algorithm which is inversely proportional to classification error, balance between cluster, number of feature used. The proposed design scheme is applied to the handwtitten alphabetic characters. Experimental results show the usefulness of the proposed scheme.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.2
• /
• pp.191-196
• /
• 2009

In this paper, we proposed a feature selection method using Binary Particle Swarm Optimization(BPSO) and Mutual information. This proposed method consists of the feature selection part for selecting candidate feature subset by mutual information and the optimal feature selection part for choosing optimal feature subset by BPSO in the candidate feature subsets. In the candidate feature selection part, we computed the mutual information of all features, respectively and selected a candidate feature subset by the ranking of mutual information. In the optimal feature selection part, optimal feature subset can be found by BPSO in the candidate feature subset. In the BPSO process, we used multi-object function to optimize both accuracy of classifier and selected feature subset size. DNA expression dataset are used for estimating the performance of the proposed method. Experimental results show that this method can achieve better performance for pattern recognition problems than conventional ones.

• The Bulletin of The Korean Astronomical Society
• /
• v.37 no.2
• /
• pp.109.2-109.2
• /
• 2012

Current microlensing follow-up observations focus on high-magnification events because of the high efficiency of planet detection. However, central perturbations of high-magnification events caused by a planet can also be produced by a very close or a very wide binary companion, and the two kinds of central perturbations are not generally distinguished without time consuming detailed modeling (a planet-binary degeneracy). Hence, it is important to resolve the planet-binary degeneracy that occurs in high-magnification events. In this paper, we investigate caustic-crossing high-magnification events caused by a planet and a wide binary companion. From this study, we find that because of the different magnification excess patterns inside the central caustics induced by the planet and the binary companion, the light curves of the caustic-crossing planetary-lensing events exhibit a feature that is discriminated from those of the caustic-crossing binary-lensing events, and the feature can be used to immediately distinguish between the planetary and binary companions.

• Journal of Institute of Control, Robotics and Systems
• /
• v.2 no.2
• /
• pp.108-114
• /
• 1996

We present proofs of the stability and convergence of Self-organizing feature map (SOFM) neural network with time-invarient learning rate and binary reinforcement function. One of the major problems in Self-organizing feature map neural network concerns with learning rate-"Kalman Filter" gain in stochsatic control field which is monotone decreasing function and converges to 0 for satisfying minimum variance property. In this paper, we show that the stability and convergence of Self-organizing feature map neural network with time-invariant learning rate. The analysis of the proposed algorithm shows that the stability and convergence is guranteed with exponentially stable and weak convergence properties as well.s as well.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.6
• /
• pp.102-110
• /
• 1996

A new design scheme of a binary decision tree is proposed. In this scheme a binary decision tree is constructed by using genetic algorithm and FCM algorithm. At each node optimal or near-optimal feature subset is selected which optimizes fitness function in genetic algorithm. The fitness function is inversely proportional to classification error, balance between cluster, number of feature used. The binary strings in genetic algorithm determine the feature subset and classification results - error, balance - form fuzzy partition matrix affect reproduction of next genratin. The proposed design scheme is applied to the tire tread patterns and handwriteen alphabetic characters. Experimental results show the usefulness of the proposed scheme.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.06a
• /
• pp.208-212
• /
• 1998

This paper suggests the method to recognize the various defect patterns of cold mill strip using binary decision tree constructed by genetic algorithm automatically. In case of classifying the complex the complex patterns with high similarity like the defect patterns of cold mill strip, the selection of the optimal feature set and the structure of recognizer is important for high recognition rate. In this paper genetic algorithm is used to select a subset of the suitable features at each node in binary decision tree. The feature subset of maximum fitness is chosen and the patterns are classified into two classes by linear decision function. After this process is repeated at each node until all the patterns are classified respectively into individual classes. In this way , binary decision tree classifier is constructed automatically. After construction binary decision tree, the final recognizer is accomplished by the learning process of neural network using a set of standard p tterns at each node. In this paper, binary decision tree classifier is applied to recognition of the defect patterns of cold mill strip and the experimental results are given to show the usefulness of the proposed scheme.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.1
• /
• pp.98-103
• /
• 2000

This paper suggests a method to recognize the various defect patterns of a cold mill strip using a binary decision tree automatically constructed by a genetic algorithm(GA). In classifying complex patterns with high similarity like the defect patterns of a cold mill stirp, the selection of an optimal feature set and an appropriate recognizer is important to achieve high recognition rate. In this paper a GA is used to select a subset of the suitable features at each node in the binary decision tree. The feature subset with maximum fitness is chosen and the patterns are classified into two classes using a linear decision function. This process is repeated at each node until all the patterns are classified into individual classes. In this way, the classifier using the binary decision tree is constructed automatically. After constructing the binary decision tree, the final recognizer is accomplished by having neural network learning sits of standard patterns at each node. In this paper, the classifier using the binary decision tree is applied to the recognition of defect patterns of a cold mill strip, and the experimental results are given to demonstrate the usefulness of the proposed scheme.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.2
• /
• pp.219-224
• /
• 2020

Binary classification has been broadly investigated in machine learning. In addition, binary classification can be easily extended to multi class problems. To successfully utilize machine learning methods for classification tasks, preprocessing and feature extraction steps are essential. These are important steps to improve their classification performances. In this paper, we propose a new learning method based on weighted least squares. In the weighted least squares, designing weights has a significant role. Due to this necessity, we also propose a new technique to obtain weights that can achieve feature transformation. Based on this weighting technique, we also propose a method to combine the learning and feature extraction processes together to perform both processes simultaneously in one step. The proposed method shows the promising performance on five UCI machine learning data sets.