• MALSORI
• /
• v.55
• /
• pp.131-142
• /
• 2005

A mismatch between the training and the test conditions often causes a drastic decrease in the performance of the speech recognition systems. In this paper, non-linear transformation techniques based on histogram equalization in the acoustic feature space are studied for reducing the mismatched condition. The purpose of histogram equalization(HEQ) is to convert the probability distribution of test speech into the probability distribution of training speech. While conventional histogram equalization methods consider only the probability distribution of a test speech, for noise-corrupted test speech, its probability distribution is also distorted. The transformation function obtained by this distorted probability distribution maybe bring about miss-transformation of feature vectors, and this causes the performance of histogram equalization to decrease. Therefore, this paper proposes a new method of calculating noise-removed probability distribution by using assumption that the CDF of noisy speech feature vectors consists of component of speech feature vectors and component of noise feature vectors, and this compensated probability distribution is used in HEQ process. In the AURORA-2 framework, the proposed method reduced the error rate by over $44\%$ in clean training condition compared to the baseline system. For multi training condition, the proposed methods are also better than the baseline system.

• Journal of the Korean School Mathematics Society
• /
• v.1 no.1
• /
• pp.59-67
• /
• 1998

Probability and statistics is an important section in mathematics which is deeply related to everyday living, natural science and social science. In spite of its importance, many students will throw away it because it becomes very harder as its step(stage) deepens and probability and statistics' relative importance is very small in Korea-SAT(the test of college entrance in Korea). Therefore, by analyzing the involvement carefully between the curriculum in the elementary, middle, high school and the text book, by studying the problem and improvement direction, it is necessary to investigate an effective teaching method. This study intends to give the students the confidence, interests, and accomplishment motive about probability and statistics field and to make a rational and creative decision-making through mathematical speculation by proposing an effective teaching method through analyzing an existing facts in school's probability and statistics field. The contents of this study are composed of four chapters. Chapter three looks into the mathematical curriculum in the elementary, middle, high school and its teaching meaning, the outline of contents, some tips on teaching and problems and presents an effective and concrete teaching method on the basis of the theoretical background in the chapter two. Chapter four is a conclusive part and gives the general improvement and intentional direction in educating the probability and statistics.

• Journal of Environmental Science International
• /
• v.28 no.1
• /
• pp.19-35
• /
• 2019

This study investigates the performance of four Bayesian methods, Random Walk Metropolis (RWM), Hit-And-Run Metropolis (HARM), Adaptive Mixture Metropolis (AMM), and Population Monte Carlo (PMC), for estimating the parameters and uncertainties of probability rainfall distribution, and the results are compared with those of conventional parameter estimation methods; namely, the Method Of Moment (MOM), Maximum Likelihood Method (MLM), and Probability Weighted Method (PWM). As a result, Bayesian methods yield similar or slightly better results in parameter estimations compared with conventional methods. In particular, PMC can reduce parameter uncertainty greatly compared with RWM, HARM, and AMM methods although the Bayesian methods produce similar results in parameter estimations. Overall, the Bayesian methods produce better accuracy for scale parameters compared with the conventional methods and this characteristic improves the accuracy of probability rainfall. Therefore, Bayesian methods can be effective tools for estimating the parameters and uncertainties of probability rainfall distribution in hydrological practices, flood risk assessment, and decision-making support.

• Journal of the Korean Society of Safety
• /
• v.18 no.1
• /
• pp.116-122
• /
• 2003

Monte Carlo method is one of the powerful engineering tools especially to solve the complex non-linear problems. The Monte Carlo method gives approximate solution to a variety of mathematical problems by performing statistical sampling experiments on a computer. One of the methods to predict the time dependent failure probability of one of the bridge components or the bridge system is a lifetime function. In this paper, FORTRAN program is developed to predict the failure probability of bridge components or bridge system by using both system reliability and lifetime function. Monte Carlo method is used to generate the parameters of the lifetime function. As a case study, the program is applied to the concrete-steel bridge to predict the failure probability.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.22 no.10
• /
• pp.49-55
• /
• 2008

This paper presents a novel evaluation method of life expectancy of power system equipment. The life expectancy means expected remaining lifetime; it can be usefully utilized to maintenance planning, equipment replacement planning, and reliability assessment. The proposed method is composed of three steps. Firstly, a cumulative probability for future years is evaluated for targeted age year. Secondly, the cumulative probability is modeled by well-blown cumulative distribution function(CDF) such as Weibull distribution. Lastly, life expectancy is evaluated as the mean value of the model. Since the model CDF is established in the proposed method, it can also evaluate the probability of equipment retirement within specific years. The developed method is applied to examples of generators of combined cycle power plants to show its effectiveness.

• Geomechanics and Engineering
• /
• v.36 no.6
• /
• pp.631-640
• /
• 2024

In the traditional slope stability analysis, ignoring the spatial variability of slope soil will lead to inaccurate analysis. In this paper, the K-L series expansion method is adopted to simulate random field of soil strength parameters. Based on Random Limit Equilibrium Method (RLEM), the influence of variation coefficient and fluctuation range on reliability of soil slope supported by micro-pile is investigated. The results show that the fluctuation ranges and the variation coefficients significantly influence the failure probability of soil slope supported by micro-pile. With the increase of fluctuation range of soil strength parameters, the mean safety factor of the slope increases slightly. The failure probability of the soil slope increases with the increase of fluctuation range when the mean safety factor of the slope is greater than 1. The failure probability of the slope increases by nearly 8.5% when the fluctuation range is increased from δ_v=2 m to δ_v =8 m. With the increase of the variation coefficient of soil strength parameters, the mean safety factor of the slope decreases slightly, and the probability of failure of soil slope increases accordingly. The failure probability of the slope increases by nearly 31% when the variation coefficient increases from COV_c=0.2, COV_φ=0.05 to COV_c=0.5, COV_φ=0.2.

• Journal of the Korean Data and Information Science Society
• /
• v.10 no.2
• /
• pp.415-428
• /
• 1999

In this paper, we study the saddlepoint approximations for the ratio of two independent sequences of random variables. In Section 2, we review the saddlepoint approximation to the probability density function. In section 3, we derive an saddlepoint approximation formular for the tail probability by following Daniels'(1987) method. In Section 4, we represent a numerical example which shows that the errors are small even for small sample size.

• The Mathematical Education
• /
• v.45 no.4 s.115
• /
• pp.381-406
• /
• 2006

This work aims to investigate mathematics teachers' knowledge and belief on the high school probability and statistics. For this aim, two research questions are estabilished as follows. (1) How is mathematics teachers' knowledge on the main contents of the high school probability and statistics in the 7th mathematics curriculum? (2) What is mathematics teachers' belief on the high school probability and statistics? Survey and interviews were carried out to answer the above research questions. Subjects of the survey were 2 7mathematics teachers who were answered to questionnaire. Among them, 3 volunteers were chosen by provinces for in-depth interview. Research findings in mathematics teacher's knowledge are as follows. Firstly, mathematics teachers do not have much of mathematical knowledge on the newly added and changed contents of the high school probability and statistics in the 7th mathematics curriculum. Secondly, mathematics teachers do not change their teaching-learning method for probability and statistics. Thirdly, many teachers think that the use of technology and reconstruction of the textbooks are required in teaching and learning of the high school probability and statistics. But, they stick on their own way. Research findings in mathematics teachers' belief are as follows. Firstly, many mathematics teachers view the nature of statistics as a branch of the applied mathematics and put the value of high school probability and statistics on the practical usefulness, Secondly, many mathematics teachers think that understanding concepts and improving problem solving ability are the best method of the teaching and learning. Thirdly, many mathematics teachers think that high school probability and statistics textbooks should cause motivations and interests in order not to give up studying probability and statistics. It is expected that the above findings can be used to change teachers' teaching and learning methods and to improve teachers training program.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.41 no.2
• /
• pp.156-163
• /
• 2013

Collision probability is the most common method to measure the risk of space debris, it is widely used that two dimensional linear collision probability using the closest approach distance. This paper represents the characteristics of object that approach KOMPSAT 2, 3, 5 that have operated or will be operated by Korea. And more precise method than two dimensional linear collision probability, we analyzed the properties of three dimensional nonlinear collision probability using STK/Nonlinear Collision Probability Tool. Through this, efficiency of three dimensional nonlinear collision probability for KOMPSAT series satellites was investigated. The result represents that three dimensional nonlinear collision probability showed the precise outcome at a relative velocity of less than 350m/s. Also, KOMPSAT series satellites appeared to few low relative velocity approaches and showed low efficiency for the three dimensional nonlinear collision probability.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.62 no.1
• /
• pp.140-142
• /
• 2013

Cognitive radio (CR) has been recently proposed to dynamically access unused-spectrum. Since the spectrum availability for opportunistic access is determined by spectrum sensing, sensing control is identified as one of the most crucial issues of cognitive radio networks. Out-of-band sensing to find an available channels to sense. Sensing is also required in case of spectrum hand-off. Sensing process needs to be done very fast in order to enhance the quality of service (QoS) of the CR nodes, and transmission not to be cut for longer time. During the sensing, the PU(primary user) detection probability condition should be satisfied. We adopt a channel prediction method to find target channels. Proposed prediction method combines chaotic global method and chaotic local method for channel idle probability prediction. Global method focus on channel history information length and order number of prediction model. Local method focus on local prediction trend. Through making simulation, Proposed method can find an available channel with very high probability, total sensing time is minimized, detection probability of PU's are satisfied.