• The Transactions of the Korea Information Processing Society
• /
• v.4 no.4
• /
• pp.963-972
• /
• 1997

In imformation retrival, query reformulation algorithms construct querise from a set of intial input and feedback documents, and retrieval performance cna be varied by different sets of input documents. In this study, we developed a criterion for measuring the performance sensitivity of query reformulation algorithms to unput sets. In addition, we also propose a way of mesuring the changes in retrived area, (CIRA) during qucry reformulation. We cimpared CIRAs of query refromulation algorithms (i.e., query tree, DNF method, and Dillon's method) using three test sets:the CACM, CISI, and Medlars. In the experiments, the query tree showed the highest decreasing CIRA during refirmulations, which means the fastest convergence rate to an output set. For sensitivity analysis, the query tree sored the highest sensitivity to different input sets even though its differences to the tther algorithms are very small.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.12
• /
• pp.2313-2321
• /
• 2008

In order to develop reliable on-site partial discharge(PD) pattern recognition algorithm, the fuzzy neural network based on fuzzy set(FNN) and the polynomial network pattern classifier based on fuzzy Inference(PNC) were investigated and designed. Using PD data measured from laboratory defect models, these algorithms were learned and tested. Considering on-site situation where it is not easy to obtain voltage phases in PRPDA(Phase Resolved Partial Discharge Analysis), the measured PD data were artificially changed with shifted voltage phases for the test of the proposed algorithms. As input vectors of the algorithms, PRPD data themselves were adopted instead of using statistical parameters such as skewness and kurtotis, to improve uncertainty of statistical parameters, even though the number of input vectors were considerably increased. Also, results of the proposed neuro-fuzzy algorithms were compared with that of conventional BP-NN(Back Propagation Neural Networks) algorithm using the same data. The FNN and PNC algorithms proposed in this study were appeared to have better performance than BP-NN algorithm.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.11
• /
• pp.1053-1059
• /
• 2013

Heading bias effects in PNS using IMUs attached to shoes are analyzed in this paper. The navigation algorithms of a single foot PNS where one IMU is attached to a foot and dual foot PNSs where two IMUs are attached to each foot are derived. Two navigation algorithms are proposed for the dual foot PNS: 1) the positions from the independent right and left foot PNSs are averaged to provide the final position, 2) the right and left foot PNSs are correlated and it provides positions of each foot. Furthermore, it is proven that two methods are equal. Using the derived navigation algorithms the effect of heading bias caused by a misalignment of the moving direction and IMU is analyzed. The analysis explains the position error of a single foot PNS is diverged while the heading bias is effectively compensated in dual foot PNSs because of the symmetry of heading biases. The experimental results confirm the analysis.

• Wind and Structures
• /
• v.11 no.6
• /
• pp.497-512
• /
• 2008

In this paper, optimum methods of wind-induced vibration control of high-rise buildings are mainly studied. Two optimum methods, genetic algorithms (GA) method and Rayleigh damping method, are firstly employed and proposed to perform optimum study on wind-induced vibration control, six target functions are presented in GA method based on spectrum analysis. Structural optimum analysis programs are developed based on Matlab software to calculate wind-induced structural responses. A high-rise steel building with 20-storey is adopted and 22 kinds of control plans are employed to perform comparison analysis to validate the feasibility and validity of the optimum methods considered. The results show that the distributions of damping coefficients along structural height for mass proportional damping (MPD) systems and stiffness proportional damping (SPD) systems are entirely opposite. Damping systems of MPD and GAMPD (genetic algorithms and mass proportional damping) have the best performance of reducing structural wind-induced vibration response and are superior to other damping systems. Standard deviations of structural responses are influenced greatly by different target functions and the influence is increasing slightly when higher modes are considered, as shown fully in section 5. Therefore, the influence of higher modes should be considered when strict requirement of wind-induced vibration comfort is needed for some special structures.

• Journal of Korean Society of Water and Wastewater
• /
• v.35 no.6
• /
• pp.507-516
• /
• 2021

In this study, the machine learning which has been widely used in prediction algorithms recently was used. the research point was the CD(chudong) point which was a representative point of Daecheong Lake. Chlorophyll-a(Chl-a) concentration was used as a target variable for algae prediction. to predict the Chl-a concentration, a data set of water quality and quantity factors was consisted. we performed algorithms about random forest and gradient boosting with Python. to perform the algorithms, at first the correlation analysis between Chl-a and water quality and quantity data was studied. we extracted ten factors of high importance for water quality and quantity data. as a result of the algorithm performance index, the gradient boosting showed that RMSE was 2.72 mg/m³ and MSE was 7.40 mg/m³ and R² was 0.66. as a result of the residual analysis, the analysis result of gradient boosting was excellent. as a result of the algorithm execution, the gradient boosting algorithm was excellent. the gradient boosting algorithm was also excellent with 2.44 mg/m³ of RMSE in the machine learning hyperparameter adjustment result.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.5
• /
• pp.989-998
• /
• 2013

We cannot analyze big data, which attracts recent attentions in industry and academy, by batch processing algorithms developed in data mining because big data, by definition, cannot be uploaded and processed in the memory of a single system. So an imminent issue is to develop various leaning algorithms so that they can be applied to big data. In this paper, we review various algorithms for support vector machines in the literature. Particularly, we introduce online type and parallel processing algorithms that are expected to be useful in big data classifications and compare the strengths, the weaknesses and the performances of those algorithms through simulations for linear classification.

• Journal of Korean Society for Quality Management
• /
• v.48 no.3
• /
• pp.421-431
• /
• 2020

Purpose: The purpose of this study is to compare and analyze the tokenizer in natural language processing for customer satisfaction in sentiment analysis. Methods: In this study, a supervised learning-based tokenizer Mecab-Ko and an unsupervised learning-based tokenizer SentencePiece were used for comparison. Three algorithms: Naïve Bayes, k-Nearest Neighbor, and Decision Tree were selected to compare the performance of each tokenizer. For performance comparison, three metrics: accuracy, precision, and recall were used in the study. Results: The results of this study are as follows; Through performance evaluation and verification, it was confirmed that SentencePiece shows better classification performance than Mecab-Ko. In order to confirm the robustness of the derived results, independent t-tests were conducted on the evaluation results for the two types of the tokenizer. As a result of the study, it was confirmed that the classification performance of the SentencePiece tokenizer was high in the k-Nearest Neighbor and Decision Tree algorithms. In addition, the Decision Tree showed slightly higher accuracy among the three classification algorithms. Conclusion: The SentencePiece tokenizer can be used to classify and interpret customer sentiment based on online reviews in Korean more accurately. In addition, it seems that it is possible to give a specific meaning to a short word or a jargon, which is often used by users when evaluating products but is not defined in advance.

• ETRI Journal
• /
• v.34 no.3
• /
• pp.478-481
• /
• 2012

Testing memory and repairing faults have become increasingly important for improving yield. Redundancy analysis (RA) algorithms have been developed to repair memory faults. However, many RA algorithms have low analysis speeds and occupy memory space within automatic test equipment. A fast RA algorithm using simple calculations is proposed in this letter to minimize both the test and repair time. This analysis uses the grouped addresses in the faulty bitmap. Since the fault groups are independent of each other, the time needed to find solutions can be greatly reduced using these fault groups. Also, the proposed algorithm does not need to store searching trees, thereby minimizing the required memory space. Our experiments show that the proposed RA algorithm is very efficient in terms of speed and memory requirements.

• Journal of The Korean Association of Information Education
• /
• v.22 no.6
• /
• pp.671-679
• /
• 2018

Sorting algorithms are the basic building blocks that computer science students need to learn. In recent years, sorting algorithms also have begun to be taught in K-12 classrooms using "the educational sorting game" described in CSUnplugged. However, although the educational sorting game was developed for students aged 8 and up, it is hard for K-12 teachers to play with their students because it is difficult for teachers to understand all of the algorithms and some popular algorithms do not work well in the educational sorting game. In this paper, we discuss what teachers should know, and experimentally analyze the performance of the existing algorithms when applied to the educational sorting game.

• The Journal of Korean Association of Computer Education
• /
• v.19 no.1
• /
• pp.101-109
• /
• 2016

Designing algorithms is a very important learning process in computational thinking education because it requires learner's logical and procedural thinking. But the case studies that have topics of algorithms learning and students' types of errors in learning algorithms are not enough. So the purpose of this study is to analyze students' errors that discovered in the process of learning three control structures of programming using flowchart and provide types of errors in designing algorithms. Results about tests of three types of control structures in university student's algorithms learning class showed different cases of types of errors; types of sequential control error are not presented in the class, types of conditional control error are presented in the case of setting the conditions of nested conditional control, and types of iterative control are showed in the many cases of iterative conditions, statements of single and nested iterative control structure. The results of study will be a good case study about teaching designing algorithms of computational thinking education in elementary, secondary school and university.