• Korean Journal of Remote Sensing
• /
• v.21 no.5
• /
• pp.417-423
• /
• 2005

A fuzzy approach using an EM algorithm for image classification is presented. In this study, a double compound stochastic image process is assumed to combine a discrete-valued field for region-class processes and a continuous random field for observed intensity processes. The Markov random field is employed to characterize the geophysical connectedness of a digital image structure. The fuzzy classification is an EM iterative approach based on mixture probability distribution. Under the assumption of the double compound process, given an initial class map, this approach iteratively computes the fuzzy membership vectors in the E-step and the estimates of class-related parameters in the M-step. In the experiments with remotely sensed data, the MRF-based method yielded a spatially smooth class-map with more distinctive configuration of the classes than the non-MRF approach.

• The Bulletin of The Korean Astronomical Society
• /
• v.45 no.1
• /
• pp.51.3-52
• /
• 2020

We present the status of the development of a Markov Chain Monte Carlo (MCMC) parameter estimation (PE) pipeline for compact binary coalescences (CBCs) with the Japanese KAGRA gravitational-wave (GW) detector. The pipeline is included in the KAGRA Algorithm Library (KAGALI). Basic functionalities are benchmarked from the LIGO Algorithm Library (LALSuite) but the KAGRA MCMC PE pipeline will provide a simpler, memory-efficient pipeline to estimate physical parameters from gravitational waves emitted from compact binaries consisting of black holes or neutron stars. Applying inspiral-merge-ringdown and inspiral waveforms, we performed simulations of various black hole binaries, we performed the code sanity check and performance test. In this talk, we present the situation of GW observation with the Covid-19 pandemic. In addition to preliminary PE results with the KAGALI MCMC PE pipeline, we discuss how we can optimize a CBC PE pipeline toward the next observation run.

• The Journal of the Acoustical Society of Korea
• /
• v.13 no.1
• /
• pp.24-31
• /
• 1994

In this paper, we report on the development of a speaker independent continuous speech recognition system using continuous hidden Markov models. The continuous hidden Markov model consists of mean and covariance matrices and directly models speech signal parameters, therefore does not have quantization error. Filter bank coefficients with their 1st and 2nd-order derivatives are used as feature vectors to represent the dynamic features of speech signal. We use the segmental K-means algorithm as a training algorithm and triphone as a recognition unit to alleviate performance degradation due to coarticulation problems critical in continuous speech recognition. Also, we use the one-pass search algorithm that Is advantageous in speeding-up the recognition time. Experimental results show that the system attains the recognition accuracy of $83\%$ without grammar and $94\%$ with finite state networks in speaker-indepdent speech recognition.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.225-225
• /
• 2015

Drought events usually evolve slowly in time and their impacts generally span a long period of time. This indicates that the sequence of drought is not completely random. The Hidden Markov Model (HMM) is a probabilistic model used to represent dependences between invisible hidden states which finally result in observations. Drought characteristics are dependent on the underlying generating mechanism, which can be well modelled by the HMM. This study employed a HMM with Gaussian emissions to fit the Standardized Precipitation Index (SPI) series and make multi-step prediction to check the drought characteristics in the future. To estimate the parameters of the HMM, we employed a Bayesian model computed via Markov Chain Monte Carlo (MCMC). Since the true number of hidden states is unknown, we fit the model with varying number of hidden states and used reversible jump to allow for transdimensional moves between models with different numbers of states. We applied the HMM to several stations SPI data in South Korea. The monthly SPI data from January 1973 to December 2012 was divided into two parts, the first 30-year SPI data (January 1973 to December 2002) was used for model calibration and the last 10-year SPI data (January 2003 to December 2012) for model validation. All the SPI data was preprocessed through the wavelet denoising and applied as the visible output in the HMM. Different lead time (T= 1, 3, 6, 12 months) forecasting performances were compared with conventional forecasting techniques (e.g., ANN and ARMA). Based on statistical evaluation performance, the HMM exhibited significant preferable results compared to conventional models with much larger forecasting skill score (about 0.3-0.6) and lower Root Mean Square Error (RMSE) values (about 0.5-0.9).

• The Korean Journal of Applied Statistics
• /
• v.36 no.1
• /
• pp.85-100
• /
• 2023

The stochastic volatility (SV) model is one of the main methods of modeling time-varying volatility. In particular, SV model is actively used in estimation and prediction of financial market volatility and option pricing. This paper attempts to model the time-varying volatility of the bitcoin market price using SV model. Hidden Markov model (HMM) is combined with the SV model to capture characteristics of regime switching of the market. The HMM is useful for recognizing patterns of time series to divide the regime of market volatility. This study estimated the volatility of bitcoin by using data from Upbit, a cryptocurrency trading site, and analyzed it by dividing the volatility regime of the market to improve the performance of the SV model. The MCMC technique is used to estimate the parameters of the SV model, and the performance of the model is verified through evaluation criteria such as MAPE and MSE.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.36 no.3
• /
• pp.203-211
• /
• 2023

A Markov chain Monte Carlo (MCMC) simulation is proposed for probabilistic full waveform inversion (FWI) in a layered half-space. Dynamic responses on the half-space surface are estimated using the thin-layer method when a harmonic vertical force is applied. Subsequently, a posterior probability distribution function and the corresponding objective function are formulated to minimize the difference between estimations and observed data as well as that of model parameters from prior information. Based on the gradient of the objective function, a proposal distribution and an acceptance probability for MCMC samples are proposed. The proposed MCMC simulation is applied to several layered half-space examples. It is demonstrated that the proposed MCMC simulation for probabilistic FWI can estimate probabilistic material properties such as the shear-wave velocities of a layered half-space.

• MALSORI
• /
• no.46
• /
• pp.37-50
• /
• 2003

In this paper, we evaluate and compare the performance of speech/music discrimination based on various feature parameters and classifiers. As for feature parameters, we consider High Zero Crossing Rate Ratio (HZCRR), Low Short Time Energy Ratio (LSTER), Spectral Flux (SF), Line Spectral Pair (LSP) distance, entropy and dynamism. We also examine three classifiers: k Nearest Neighbor (k-NN), Gaussian Mixure Model (GMM), and Hidden Markov Model (HMM). According to our experiments, LSP distance and phoneme-recognizer-based feature set (entropy and dunamism) show good performance, while performance differences due to different classifiers are not significant. When all the six feature parameters are employed, average speech/music discrimination accuracy up to 96.6% is achieved.

• The Korean Journal of Applied Statistics
• /
• v.27 no.6
• /
• pp.909-922
• /
• 2014

This paper reviews Bayesian inference with inequality constraints. It focuses on ⅰ) comparison of models with various inequality/equality constraints on parameters, ⅱ) multiple tests on equalities of parameters when parameters are under inequality constraints, ⅲ) multiple test on equalities of score parameters in models for contingency tables with ordinal categorical variables.

• The Journal of the Korea Contents Association
• /
• v.16 no.10
• /
• pp.299-310
• /
• 2016

In this paper, we perform analysis of the performance of low-power communication technology, IEEE 802.15.4 MAC (Media Access Control) protocol used in internet of things. Wide variety of devices in internet of things generate different kinds of traffic characteristics and heterogeneous traffic environment is therefore considered in contrast to existing works. With the help of the Markov chain the performance of the protocol is derived as a function of the MAC parameters and based on the results a framework to determine energy-efficient optimal MAC parameters for applications in real time is proposed. Proposed method is shown to be more efficient in terms of energy consumption through performance comparison over application scenarios compared to operation with fixing MAC parameters, and its practicability on small devices with limited computing power is verified through evaluating the accuracy and speed of the proposed framework.

• 한국데이터정보과학회:학술대회논문집
• /
• 2006.11a
• /
• pp.71-80
• /
• 2006

A nonparametric Bayesian multiple comparisons problem (MCP) for dependence parameters in I bivariate exponential populations is studied here. A simple method for pairwise comparisons of these parameters is also suggested. Here we extend the methodology studied by Gopalan and Berry (1998) using Dirichlet process priors. The family of Dirichlet process priors is applied in the form of baseline prior and likelihood combination to provide the comparisons. Computation of the posterior probabilities of all possible hypotheses are carried out through Markov Chain Monte Carlo method, namely, Gibbs sampling, due to the intractability of analytic evaluation. The whole process of MCP for the dependent parameters of bivariate exponential populations is illustrated through a numerical example.