• Proceedings of the KIEE Conference
• /
• 1998.07f
• /
• pp.1937-1939
• /
• 1998

This paper presents a simple switching method to generate a PWM pattern mostly relevant to signle-phase three-level PWM rectifier. The adopted PWM switching pattern is performed in a manner similar to the space vector PWM method, which is popularly used in the three-phase rectifier and inverter. A set of possible voltages has been selected so that an equation with a time integral considered within a sampling period should be satisfied every sampling time. The simulation result shows that the proposed control scheme is good in some performance criteria such as unity power factor, low harmonic distortion of input current, dynamic response and voltage balancing of two series-connected DC capacitors.

• Journal of Korean Institute of Industrial Engineers
• /
• v.36 no.1
• /
• pp.13-21
• /
• 2010

The prediction of contract cancellation of customers is essential in insurance companies but it is a difficult problem because the customer database is large and the target or cancelled customers are a small proportion of the database. This paper proposes a new data mining approach to the binary classification by handling a large-scale unbalanced data. Over-sampling, clustering, regularized logistic regression and boosting are also incorporated in the proposed approach. The proposed approach was applied to a real data set in the area of insurance and the results were compared with some other classification techniques.

• Journal of Korea Multimedia Society
• /
• v.18 no.7
• /
• pp.843-852
• /
• 2015

Binarization of document images is a critical pre-processing step required for character recognition. Even though various research efforts have been devoted, the quality of binarization results largely depends on the noise amount and condition of images. We propose a new binarization method that combines Maximally Stable External Region(MSER) with down-sampling. Particularly, we propose to apply different threshold values for character regions, which turns out to be effective in reducing noise. Through a set of experiments on test images, we confirmed that the proposed method was superior to existing methods in reducing noise, while the increase of execution time is limited.

• Journal of Korea Multimedia Society
• /
• v.23 no.4
• /
• pp.595-602
• /
• 2020

We propose a generative probabilistic model with Dirichlet prior distribution for topic modeling and text similarity analysis. It assigns a topic and calculates text correlation between documents within a corpus. It also provides posterior probabilities that are assigned to each topic of a document based on the prior distribution in the corpus. We then present a Gibbs sampling algorithm for inference about the posterior distribution and compute text correlation among 50 abstracts from the papers published by IEEE. We also conduct a supervised learning to set a benchmark that justifies the performance of the LDA (Latent Dirichlet Allocation). The experiments show that the accuracy for topic assignment to a certain document is 76% for LDA. The results for supervised learning show the accuracy of 61%, the precision of 93% and the f1-score of 96%. A discussion for experimental results indicates a thorough justification based on probabilities, distributions, evaluation metrics and correlation coefficients with respect to topic assignment.

• The Magazine of the IEIE
• /
• v.38 no.1
• /
• pp.56-67
• /
• 2011

As a result of quickly growing data, a digital transmission system is required to deal with the challenge of acquiring signals at a very high sampling rate, Fortunately, the CS (Compressed Sensing or Compressive Sensing) theory, a new concept based on theoretical results of signal reconstruction, can be employed to exploit the sparsity of the received signals. Then, they can be adequately reconstructed from a set of their random projections, leading to dramatic reduction in the sampling rate and in the use of ADC (Analog-to-Digital Converter) resources. The goal of this article is provide an overview of the basic CS theory and to survey some important compressed sensing applications in wireless communications.

• Journal of the Korean Data and Information Science Society
• /
• v.10 no.1
• /
• pp.119-133
• /
• 1999

A Markov Chain Monte Carlo method is developed to compute the software reliability model. We consider computation problem for determining of posterior distibution in Bayseian inference. Metropolis algorithms along with Gibbs sampling are proposed to preform the Bayesian inference of the Mixed model with record value statistics. For model determiniation, we explored the prequential conditional predictive ordinate criterion that selects the best model with the largest posterior likelihood among models using all possible subsets of the component intensity functions. To relax the monotonic intensity function assumptions. A numerical example with simulated data set is given.

• Journal of the Korea Safety Management & Science
• /
• v.9 no.2
• /
• pp.215-233
• /
• 2007

Even though the ad hoc Shewhart methods remain controversial due to various mathematical flaws, there is little disagreement among researchers and practitioners when a set of process data has a skewness distribution. In the context and language of process control, the error related to the process data shows that time to signal increases when a control parameter shifts to a skewness direction. In real-world industrial settings, however, quality practitioners often need to consider a skewness distribution. To address this situation, we developed an enhanced design method to utilize advantages of the traditional attribute control chart and to overcome its associated shortcomings. The proposed design method minimizes bias, i.e., an average time to signal for the shift of process from the target value (ATS) curve, as well as it applies a variable sampling interval (VSI) method to an attribute control chart for detecting a process shift efficiently. The results of the factorial experiment obtained by various parameter circumstances show that the VSI c control chart using nearly unbiased ATS design provides the smallest decreasing rate in ATS among other charts for all experimental cases.

• Proceedings of the KIEE Conference
• /
• 1999.07g
• /
• pp.3147-3149
• /
• 1999

In this paper, we develop a TMS320C31- 60 DSP board to generate spiral gradient waveforms for Spiral imaging, one of the ultra fast MRI methods. In Spiral imaging, accurate gradient waveforms are very important to acquire high quality image. For this purpose, sampling rate for generating the gradient waveforms is set twice as high as the data sampling rate. With the developed DSP board accurate gradient waveforms are obtained. Ultra fast MR image with the developed DSP board is currently under development.

• The Journal of Asian Finance, Economics and Business
• /
• v.7 no.6
• /
• pp.29-37
• /
• 2020

This research seeks to determine the influence of investment opportunity set (IOS); profitability (Return on Assets - ROA), liquidity, business risk and firm size on debt policy. We used 42 manufacturing companies registered on the Indonesian Stock Exchange (Bursa Efek Indonesia) as object research. We used purposive sampling method to determined samples, consider the period observation from 2012 to 2016, and produce 168 units analysis. Data analysis uses the multiple regressions with the SPSS tools. The results of the study found that companies' debt policies in Indonesia are negatively affected by the liquidity. Investment opportunity set (IOS) has negative effect on debt policy. Meanwhile, ROA, Return on Invested Capital (ROIC), and firm size of a company has no impact on debt policy. These findings indicate that Indonesian manufacture companies do not see the high investment opportunity set and profitability as a policy basis for increasing debt. Moreover, the high profitability also does not cause companies to increase their debt ratio. Our study indicates that Indonesian manufacture companies use internal funds to fund their investment. This finding is a concern for creditors, as they can now see the ability of the companies, and especially their performance, in determining their credit policies.

• Journal of Korean Society for Atmospheric Environment
• /
• v.9 no.4
• /
• pp.303-309
• /
• 1993

An observational study of urban heat island was carried out using field data obatined during 6 days in May and August 1992 in Chunchon(population size 180.000). Air temperature was measured at 64 points along two sampling ruoutes by themisters attached to cars. Both routes cover urban and rural area and across the cneter of urban area. Continuous observation of air sonde was perfomed to clarify heights of nocturnal boundary layer(NBL) at the center of urban area. Surface meteorological observations were performed at both urban and rural sites. This study showed that heat island phenomena was obviously observed at the urbanized area during the night time with low wind speed. The average NBL heights exteded to about 10 meters, but varied with meteorological conditions. After sunset, the air temperature decreased with time at both sites and cooling rate at the urban site was greater than the rural site. The maximum heat island intensity was 7.5$^{\circ}$C at 21 LST, May 4. Usingthe two meteorological data sets obtained from urban and rural sites, the air pollutant concentration was calculated by Gaussian plume model which can obtain not only horizontal distribution of concentration but also vertical distribution. The result indicated that the concentration resulted from urban meteorological data set was lower than that from rural meteorological data set. It was also calculated that the air pollutant extended to higher level in urban meteorological data set than that in rural meteorological data set.