• 대한음성학회지:말소리
• /
• 제46호
• /
• pp.117-126
• /
• 2003

This paper presents our language model adaptation for Korean spontaneous speech recognition. Korean spontaneous speech is observed various characteristics of content and style such as filled pauses, word omission, and contraction as compared with the written text corpus. Our approaches focus on improving the estimation of domain-dependent n-gram models by relevance weighting out-of-domain text data, where style is represented by n-gram based tf/sup */idf similarity. In addition to relevance weighting, we use disfluencies as Predictor to the neighboring words. The best result reduces 9.7% word error rate relatively and shows that n-gram based relevance weighting reflects style difference greatly and disfluencies are good predictor also.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2002년도 추계학술대회논문집
• /
• pp.295-300
• /
• 2002

In conventional model updating using single-objective optimization techniques, incompatible physical data are compared with each other using weighting factors. There are no general rules for selecting the weighting factors since they are not directly related with the dynamic behavior of an updated model. So one of the most difficult tasks, in model updating study, is 'balancing among the correlations' i.e. 'trade-off'. In this work, a multiobjecitive optimization technique called 'satisficing trade-off method' is introduced to extremize several correlations simultaneously. The absurd need for the weighting factors can be avoided using this technique. And the updated model with the most appropriate correlations is obtained easily in interactive way. Especially automatic trade-off is employed to increase the rate of convergence to the desired model. Its effectiveness is verified by application to a real engineering problem, HDD cover model updating.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2003년도 5월 학술대회지
• /
• pp.83-86
• /
• 2003

This paper presents our style-based language model adaptation for Korean conversational speech recognition. Korean conversational speech is observed various characteristics of content and style such as filled pauses, word omission, and contraction as compared with the written text corpora. For style-based language model adaptation, we report two approaches. Our approaches focus on improving the estimation of domain-dependent n-gram models by relevance weighting out-of-domain text data, where style is represented by n-gram based tf*idf similarity. In addition to relevance weighting, we use disfluencies as predictor to the neighboring words. The best result reduces 6.5％ word error rate absolutely and shows that n-gram based relevance weighting reflects style difference greatly and disfluencies are good predictor.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제5권11호
• /
• pp.2068-2086
• /
• 2011

Based on the quincunx sub-sampling grid, the New Interleaved Hierarchical INTerpolation (NIHINT) method is recognized as a superior pyramid data structure for the lossless and progressive coding of natural images. In this paper, we propose a new image interpolation algorithm, Edge Adaptive Hierarchical INTerpolation (EAHINT), for a further reduction in the entropy of interpolation errors. We compute the local variance of the causal context to model the strength of a local edge around a target pixel and then apply three statistical decision rules to classify the local edge into a strong edge, a weak edge, or a medium edge. According to these local edge types, we apply an interpolation method to the target pixel using a one-directional interpolator for a strong edge, a multi-directional adaptive weighting interpolator for a medium edge, or a non-directional static weighting linear interpolator for a weak edge. Experimental results show that the proposed algorithm achieves a better compression bit rate than the NIHINT method for lossless image coding. It is shown that the compression bit rate is much better for images that are rich in directional edges and textures. Our algorithm also shows better rate-distortion performance and visual quality for progressive image transmission.

• 대한전기학회논문지:전력기술부문A
• /
• 제55권2호
• /
• pp.67-69
• /
• 2006

A load composition rate estimation algorithm is developed for DSM evaluation system. The algorithm has the structure which is composed of data verification and development and can enhance the data quality Also a hourly weighting function is introduced for maintaining load shapes. The load composition rates of specific industrial customers are obtained and the results of case studies show that a reasonable load composition rate is achieved. Additionally qualify function deployment (QFD) is introduced to enhance quality and reliability of data.

• 상하수도학회지
• /
• 제16권6호
• /
• pp.710-716
• /
• 2002

A General Pipe Break Prediction Model that incorporates linear and exponential models in its form is developed. The model is capable of fitting pipe break trends that have linear, exponential or in between of linear and exponential trend by using a weighting factor. The weighting factor is adjusted to obtain a best model that minimizes the sum of squared errors of the model. The model essentially plots a best curve (or a line) passing through "cumulative number of pipe breaks" versus "break times since installation of a pipe" data points. Therefore, it prevents over-predicting future number of pipe breaks compared to the conventional exponential model. The optimal replacement time equation is derived by using the Threshold Break Rate equation by Loganathan et al. (2002).

• 유체기계공업학회:학술대회논문집
• /
• 유체기계공업학회 2005년도 연구개발 발표회 논문집
• /
• pp.185-190
• /
• 2005

This work presents a numerical procedure to optimize the elliptic-shaped pin fin arrays to enhance turbulent heat transfer. The response surface method is used as an optimization technique with Reynolds-averaged Navier Stokes analysis of flow and heat transfer. Shear stress transport (SST) turbulence model is used as a turbulence closure. Computational results for average heat transfer rate show a reasonable agreement with the experimental data. Four variables including major axis length, minor axis length, pitch and the pin fin length nondimensionalized by duct height are chosen as design variables. The objective function is defined as a linear combination of heat transfer and friction-loss related terms with weighting factor. D-optimal design is used to reduce the data points, and, with only 28 points, reliable response surface is obtained. Optimum shapes of the pin-fin arrays have been obtained in the range from 0.0 to 0.1 of weighting factor.

• 경영과학
• /
• 제11권3호
• /
• pp.153-167
• /
• 1994

Credibility theory has provided with a useful tool the assignment of weighting factor that reflects the credibility of the observed individual and collective experience to secure fair experience rate-,making. We review credibility models which can effectively estimate risk premiums using credibility theory, and suggest an empirical Bayed model based on the collective statistics to estimate the structural parameters. To illustrate the use of evolutionary models, the models are applied to the actual data, such as loss ratio, claim frequencies and severity, in the Korean automobile insurance. Also the possibilities of generalizations and applications of empirical models are discussed.

• 대한원격탐사학회지
• /
• 제27권6호
• /
• pp.653-662
• /
• 2011

Land surface temperature (LST) is a one of the key variables of land surface which can be estimated from geostationary meteorological satellite. In this study, we have developed the three sets of LST retrieval algorithm from MTSAT-2 data through the radiative transfer simulations under various atmospheric profiles (TIGR data), satellite zenith angle, spectral emissivity, and surface lapse rate conditions using MODTRAN 4. The three LST algorithms are daytime, nighttime and total LST algorithms. The weighting method based on the solar zenith angle is developed for the consistent retrieval of LST at the early morning and evening time. The spectral emissivity of two thermal infrared channels is estimated by using vegetation coverage method with land cover map and 15-day normalized vegetation index data. In general, the three LST algorithms well estimated the LST without regard to the satellite zenith angle, water vapour amount, and surface lapse rate. However, the daytime LST algorithm shows a large bias especially for the warm LST (> 300 K) at day time conditions. The night LST algorithm shows a relatively large error for the LST (260 ~ 280K) at the night time conditions. The sensitivity analysis showed that the performance of weighting method is clearly improved regardless of the impacting conditions although the improvements of the weighted LST compared to the total LST are quite different according to the atmospheric and surface lapse rate conditions. The validation results of daytime (nighttime) LST with MODIS LST showed that the correlation coefficients, bias and RMSE are about 0.62~0.93 (0.44~0.83), -1.47~1.53 (-1.80~0.17), and 2.25~4.77 (2.15~4.27), respectively. However, the performance of daytime/nighttime LST algorithms is slightly degraded compared to that of the total LST algorithm.

• Journal of Electrical Engineering and Technology
• /
• 제4권1호
• /
• pp.43-48
• /
• 2009

The purpose of the this paper is to make decision of the maintenance priority of power distribution system using Time-Varying Failure Rate(TVFR) with interruption cost. This paper emphasizes the practical use of the reliability indices and interruption cost. To make a decision of maintenance priority on power distribution system equipment, the quantification of the reliability level should be represented as a cost. In this paper, the TVFR of power distribution system equipment applied in this paper utilizes analytic method to use the historical data of KEPCO. From this result, the sensitivity analysis on TVFR of equipment was done for the priority, which represents that high priority of the equipment has more effect on system reliability, such as SAIDI or SAIFI, than other equipment. By this priority, the investment plan is established. In this result, customer interruption cost(CIC) could be extracted, and CIC is used as weighting factor to consider a importance of customer. After that, the result calculated the proposal method in this paper is compared with other priority method, such as lifetime, failure rate or only sensitivity.