• 제어로봇시스템학회:학술대회논문집
• /
• 1987.10b
• /
• pp.150-154
• /
• 1987

A new mixed approximation method is proposed for the model reduction of high order linear and time-invariant dynamic systems. This method makes allowance for stability and feature retention simultaneously. After defining dominant frequency range which affects relative stability of systems, a part of denominator is obtained using the energy dispersion method and tests are obtained using dominant frequency response matching method. The proposed method reflects the characteristic of the original system more faithfully and guarantees absolute stability of the reduction model.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.1272-1274
• /
• 2003

In this paper, we present a new algorithm for high dimensional data reduction based on wavelet decomposition and Genetic Algorithm (GA). Comparative results show the superiority of our algorithm for dimensionality reduction and accuracy improvement.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.4 s.310
• /
• pp.87-95
• /
• 2006

In this paper, we propose a novel feature extraction method for face recognition, based on Discrete Cosine Transform (DCT), Energy Probability (EP), and Linear Discriminant Analysis (LDA). We define an energy probability as magnitude of effective information and it is used to create a frequency mask in OCT domain. The feature extraction method consists of three steps; i) the spatial domain of face images is transformed into the frequency domain called OCT domain; ii) energy property is applied on DCT domain that acquire from face image for the purpose of dimension reduction of data and optimization of valid information; iii) in order to obtain the most significant and invariant feature of face images, LDA is applied to the data extracted using frequency mask. In experiments, the recognition rate is 96.8% in ETRI database and 100% in ORL database. The proposed method has been shown improvements on the dimension reduction of feature space and the face recognition over the previously proposed methods.

• Journal of Internet Computing and Services
• /
• v.21 no.4
• /
• pp.127-135
• /
• 2020

The purpose of this study is to develop a human activity recognition (HAR) system, which distinguishes 13 activities, including five activities commonly dealt with in conventional HAR researches and eight activities from the Eastern-Asian culture. The eight special activities include floor-sitting/standing, chair-sitting/standing, floor-lying/up, and bed-lying/up. We used a 3-axis accelerometer sensor on the wrist for data collection and designed a machine learning model for the activity classification. Data clustering through preprocessing and feature extraction/reduction is performed. We then tested six machine learning algorithms for recognition accuracy comparison. As a result, we have achieved an average accuracy of 99.7% for the 13 activities. This result is far better than the average accuracy of current HAR researches based on a smartwatch (89.4%). The superiority of the HAR system developed in this study is proven because we have achieved 98.7% accuracy with publically available 'pamap2' dataset of 12 activities, whose conventionally met the best accuracy is 96.6%.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.64 no.7
• /
• pp.1064-1073
• /
• 2015

In this study, we introduce the pedestrian detection system by using the feature of HOG-PCA and RBFNNs pattern classifier. HOG(Histogram of Oriented Gradient) feature is extracted from input image to identify and recognize a object. And a dimension is reduced for improving performance as well as processing speed by using PCA which is a typical dimensional reduction algorithm. So, the feature of HOG-PCA through the dimensional reduction by using PCA leads to the improvement of the detection rate. FCM clustering algorithm is used instead of gaussian function to apply the characteristic of input data as well and connection weight is used by polynomial expression such as constant, linear, quadratic and modified quadratic. Finally, INRIA person database known as one of the benchmark dataset used for pedestrian detection is applied for the performance evaluation of the proposed classifier. The experimental result of the proposed classifier are compared with those studied by Dalal.

• The Journal of the Acoustical Society of Korea
• /
• v.37 no.6
• /
• pp.518-524
• /
• 2018

The cover song refers to live recordings or reproduced albums. This paper studies two-dimensional Fourier transform as a feature-dimension reduction method to search cover song fast. The two-dimensional Fourier transform is conducive in feature-dimension reduction for cover song search due to musical-key invariance. This paper extends the previous work, which only utilize the magnitude of the Fourier transform, by introducing an invariant from phase based on the assumption that adjacent frames have the same musical-key change. We compare the cover song retrieval accuracy of the Fourier-transform based methods over two datasets. The experimental results show that the addition of the invariant from phase improves the cover song retrieval accuracy over the previous magnitude-only method.

• Journal of Internet Computing and Services
• /
• v.22 no.3
• /
• pp.9-16
• /
• 2021

This study aims to develop a human activity recognition (HAR) system as a Deep-Learning (DL) classification model, distinguishing various human activities. We solely rely on the signals from a wristband accelerometer worn by a person for the user's convenience. 3-axis sequential acceleration signal data are gathered within a predefined time-window-slice, and they are used as input to the classification system. We are particularly interested in developing a Deep-Learning model that can outperform conventional machine learning classification performance. A total of 13 activities based on the laboratory experiments' data are used for the initial performance comparison. We have improved classification performance using the Convolutional Neural Network (CNN) combined with an auto-encoder feature reduction and parameter tuning. With various publically available HAR datasets, we could also achieve significant improvement in HAR classification. Our CNN model is also compared against Recurrent-Neural-Network(RNN) with Long Short-Term Memory(LSTM) to demonstrate its superiority. Noticeably, our model could distinguish both general activities and near-identical activities such as sitting down on the chair and floor, with almost perfect classification accuracy.

• International Journal of Computer Science & Network Security
• /
• v.22 no.10
• /
• pp.1-10
• /
• 2022

Nowadays software defect prediction (SDP) is most active research going on in software engineering. Early detection of defects lowers the cost of the software and also improves reliability. Machine learning techniques are widely used to create SDP models based on programming measures. The majority of defect prediction models in the literature have problems with class imbalance and high dimensionality. In this paper, we proposed Centroid and Nearest Neighbor based Class Imbalance Reduction (CNNCIR) technique that considers dataset distribution characteristics to generate symmetry between defective and non-defective records in imbalanced datasets. The proposed approach is compared with SMOTE (Synthetic Minority Oversampling Technique). The high-dimensionality problem is addressed using Ant Colony Optimization (ACO) technique by choosing relevant features. We used nine different classifiers to analyze six open-source software defect datasets from the PROMISE repository and seven performance measures are used to evaluate them. The results of the proposed CNNCIR method with ACO based feature selection reveals that it outperforms SMOTE in the majority of cases.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.504-509
• /
• 2008

Support Vector Machine (SVM) is one of powerful learning machine and has been applied to varying task with generally acceptable performance. The success of SVM for classification tasks in one domain is affected by features which represent the instance of specific class. Given the representative and discriminative features, SVM learning will give good generalization and consequently we can obtain good classifier. In this paper, we will assess the problem of feature choices for human detection tasks and measure the performance of each feature. Here we will consider HOG-family feature. As a natural extension of SVM, we combine SVM with Principal Component Analysis (PCA) to reduce dimension of features while retaining most of discriminative feature vectors.

• Convergence Security Journal
• /
• v.6 no.3
• /
• pp.87-95
• /
• 2006

An intrusion can be defined as any set of actors that attempt to compromise the integrity, confidentiality and availability of computer resource and destroy the security policy of computer system. The Intrusion Detection System that detects the intrusion consists of data collection, data reduction, analysis and detection, and report and response. It is important for feature selection to detect the intrusion efficiently after collecting the large set of data of Intrusion Detection System. In this paper, the feature selection method using Genetic Algorithm and Decision Tree is proposed. Also the method is verified by the simulation with KDD data.