• The KIPS Transactions:PartB
• /
• v.14B no.7
• /
• pp.473-480
• /
• 2007

Feature tracking in video frame sequences has suffered from the instability and the frequent failure of feature matching between two successive frames. In this paper, we propose a robust 2D feature tracking method that is stable to long video sequences. To improve the stability of feature tracking, we predict the spatial movement in the current image frame using the state variables. The predicted current movement is used for the initialization of the search window. By computing the feature similarities in the search window, we refine the current feature positions. Then, the current feature states are updated. This tracking process is repeated for each input frame. To reduce false matches, the outlier rejection stage is also introduced. Experimental results from real video sequences showed that the proposed method performs stable feature tracking for long frame sequences.

• Journal of Advanced Navigation Technology
• /
• v.12 no.4
• /
• pp.372-377
• /
• 2008

This paper proposes a new query-based document summarization method using the semantic feature matrix and the semantic variable matrix. The proposed method doesn't need the training phase using training data comprising queries and query specific documents. And it exactly summarizes documents for the given query by using semantic features and semantic variables that is better at identifying sub-topics of document. Because the NMF have a great power to naturally extract semantic features representing the inherent structure of a document. The experimental results show that the proposed method achieves better performance than other methods.

• ETRI Journal
• /
• v.44 no.3
• /
• pp.462-475
• /
• 2022

Although researchers have proposed numerous techniques for speech emotion recognition, its performance remains unsatisfactory in many application scenarios. In this study, we propose a speech emotion recognition model based on a genetic algorithm (GA)-decision tree (DT) fusion of deep and acoustic features. To more comprehensively express speech emotional information, first, frame-level deep and acoustic features are extracted from a speech signal. Next, five kinds of statistic variables of these features are calculated to obtain utterance-level features. The Fisher feature selection criterion is employed to select high-performance features, removing redundant information. In the feature fusion stage, the GA is is used to adaptively search for the best feature fusion weight. Finally, using the fused feature, the proposed speech emotion recognition model based on a DT support vector machine model is realized. Experimental results on the Berlin speech emotion database and the Chinese emotion speech database indicate that the proposed model outperforms an average weight fusion method.

• Journal of the Korea Safety Management & Science
• /
• v.8 no.2
• /
• 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.

• IE interfaces
• /
• v.16 no.spc
• /
• pp.39-44
• /
• 2003

As one of many design variables, the role of dimension tolerances is to restrict the amount of size variation in a manufactured feature while ensuring functionality. In this study, a nonlinear integer model has been modeled to allocate the optimal tolerance to each individual feature at a minimum manufacturing cost. While a normal distribution determines statistically worst tolerances with its symmetrical property in many previous tolerance allocation studies, a asymmetrical distribution is more realistic because its mean is not always coincident with a process center. A nonlinear integer model is modeled to allocate the optimal tolerance to a feature based on a beta distribution at a minimum total cost. The total cost as a function of tolerances is defined by machining cost and quality loss. After the convexity of manufacturing cost is checked by the Hessian matrix, the model is solved by the Complex Method. Finally, a numerical example is presented demonstrating successful model implementation for a nonlinear design case.

• Journal of Information Technology Services
• /
• v.18 no.2
• /
• pp.143-159
• /
• 2019

Machine learning (ML) is a method of fitting given data to a mathematical model to derive insights or to predict. In the age of big data, where the amount of available data increases exponentially due to the development of information technology and smart devices, ML shows high prediction performance due to pattern detection without bias. The feature engineering that generates the features that can explain the problem to be solved in the ML process has a great influence on the performance and its importance is continuously emphasized. Despite this importance, however, it is still considered a difficult task as it requires a thorough understanding of the domain characteristics as well as an understanding of source data and the iterative procedure. Therefore, we propose methods to apply deep learning for solving the complexity and difficulty of feature extraction and improving the performance of ML model. Unlike other techniques, the most common reason for the superior performance of deep learning techniques in complex unstructured data processing is that it is possible to extract features from the source data itself. In order to apply these advantages to the business problems, we propose deep learning based methods that can automatically extract features from transaction data or directly predict and classify target variables. In particular, we applied techniques that show high performance in existing text processing based on the structural similarity between transaction data and text data. And we also verified the suitability of each method according to the characteristics of transaction data. Through our study, it is possible not only to search for the possibility of automated feature extraction but also to obtain a benchmark model that shows a certain level of performance before performing the feature extraction task by a human. In addition, it is expected that it will be able to provide guidelines for choosing a suitable deep learning model based on the business problem and the data characteristics.

• International Journal of Computer Science & Network Security
• /
• v.23 no.5
• /
• pp.148-162
• /
• 2023

Classification systems can significantly assist the medical sector by allowing for the precise and quick diagnosis of diseases. As a result, both doctors and patients will save time. A possible way for identifying risk variables is to use machine learning algorithms. Non-surgical technologies, such as machine learning, are trustworthy and effective in categorizing healthy and heart-disease patients, and they save time and effort. The goal of this study is to create a medical intelligent decision support system based on machine learning for the diagnosis of heart disease. We have used a mixed feature creation (MFC) technique to generate new features from the UCI Cleveland Cardiology dataset. We select the most suitable features by using Least Absolute Shrinkage and Selection Operator (LASSO), Recursive Feature Elimination with Random Forest feature selection (RFE-RF) and the best features of both LASSO RFE-RF (BLR) techniques. Cross-validated and grid-search methods are used to optimize the parameters of the estimator used in applying these algorithms. and classifier performance assessment metrics including classification accuracy, specificity, sensitivity, precision, and F1-Score, of each classification model, along with execution time and RMSE the results are presented independently for comparison. Our proposed work finds the best potential outcome across all available prediction models and improves the system's performance, allowing physicians to diagnose heart patients more accurately.

• Phonetics and Speech Sciences
• /
• v.1 no.3
• /
• pp.19-28
• /
• 2009

This paper reports experimental findings about the duration feature of the acoustic components of English stops in Korean speakers' voicing perception. In our experiment, 35 participants discriminated between recorded stimuli and digitally transformed stimuli with different duration features from the original stimuli. 72 sets of paired stimuli are generated to test the effects of the duration feature in various phonetic contexts. The result of our experiment is a complicated cross-tabulation with 540 cells defined by five categorical independent variables plus one response variable. To find a meaningful generalization out of this complex frequency table, we ran logit log-linear regression analyses. Surprisingly, we have found that there is no single effect of the duration feature in all phonetic contexts on Korean speakers' perception of the voicing contrasts of English stops. Instead, the logit log-linear analyses reveal that there are interaction effects among phonetic contexts (=C), the places of articulation of stops (=P), and the voicing contrast (=V), and among duration (=T), phonetic contexts, and the places of articulation. To put it in mathematical terms, the distribution of the data can be explained by a simple log-linear equation, logF=${\mu}+{\lambda}CPV+{\lambda}TCP$.

• Journal of the Korean Society of Safety
• /
• v.20 no.4 s.72
• /
• pp.130-137
• /
• 2005

Decision tree algorithm is one of the data mining techniques, which conducts grouping or prediction into several sub-groups from interested groups. This technique can analyze a feature of type on groups and can be used to detect differences in the type of industrial accidents. This paper uses C4.5 algorithm for the feature analysis. The data set consists of 24,887 features through data selection from total data of 25,159 taken from 2 year observation of industrial accidents in Korea For the purpose of this paper, one target value and eight independent variables are detailed by type of industrial accidents. There are 222 total tree nodes and 151 leaf nodes after grouping. This paper Provides an acceptable level of accuracy(%) and error rate(%) in order to measure tree accuracy about created trees. The objective of this paper is to analyze the efficiency of the C4.5 algorithm to classify types of industrial accidents data and thereby identify potential weak points in disaster risk grouping.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.11
• /
• pp.541-546
• /
• 2021

Recently, there is a trend of developing various identification and prediction models for hypertension using clinical information based on artificial intelligence and machine learning around the world. However, most previous studies on identification or prediction models of hypertension lack the consideration of the ideas of non-invasive and cost-effective variables, race, region, and countries. Therefore, the objective of this study is to present hypertension prediction model that is easily understood using only general and simple sociodemographic variables. Data used in this study was based on the Korea National Health and Nutrition Examination Survey (2018). In men, the model using the naive Bayes with the wrapper-based feature subset selection method showed the highest predictive performance (ROC = 0.790, kappa = 0.396). In women, the model using the naive Bayes with correlation-based feature subset selection method showed the strongest predictive performance (ROC = 0.850, kappa = 0.495). We found that the predictive performance of hypertension based on only sociodemographic variables was higher in women than in men. We think that our models based on machine leaning may be readily used in the field of public health and epidemiology in the future because of the use of simple sociodemographic characteristics.