• Phonetics and Speech Sciences
• /
• v.13 no.4
• /
• pp.47-53
• /
• 2021

Autoregressive Text-to-Speech (TTS) models suffer from inference instability and slow inference speed. Inference instability occurs when a poorly predicted sample at time step t affects all the subsequent predictions. Slow inference speed arises from a model structure that forces the predicted samples from time steps 1 to t-1 to predict the sample at time step t. In this study, an end-to-end non-autoregressive fast text-to-speech model is suggested as a solution to these problems. The results of this study show that this model's Mean Opinion Score (MOS) is close to that of Tacotron 2 - WaveNet, while this model's inference speed and stability are higher than those of Tacotron 2 - WaveNet. Further, this study aims to offer insight into the improvement of non-autoregressive models.

• Journal of Korean Institute of Industrial Engineers
• /
• v.39 no.4
• /
• pp.247-252
• /
• 2013

Korea gas corporation (KOGAS) is responsible for the whole sale of natural gas in the domestic market. It is important to forecast the daily demand of city gas for supply and demand control, and delivery management. Since there is the autoregressive characteristic in the daily gas demand, we introduce a modified autoregressive model as the first step. The daily gas demand also has a close connection with the outdoor temperature. Accordingly, our second proposed model is a temperature-based model. Those two models, however, do not meet the requirement for forecasting performances. To produce acceptable forecasting performances, we develop a weighted average model which compounds the autoregressive model and the temperature model. To examine our proposed methods, the forecasting results are provided. We confirm that our method can forecast the daily city gas demand accurately with reasonable performances.

• Korea Real Estate Review
• /
• v.24 no.1
• /
• pp.7-14
• /
• 2014

This study, as the temporal and spatial data for the real price apartment in Seoul from January 2006 to June 2013, empirically compared and analyzed the estimation result of apartment price using OLS by hedonic price model for the problem of space-time correlation, temporal autoregressive model (TAR) considering temporal effect, spatial autoregressive model (SAR) spatial effect and spatiotemporal autoregressive model (STAR) spatiotemporal effect. As a result, the adjusted R-square of STAR model was increased by 10% compared that of OLS model while the root mean squares error (RMSE) was decreased by 18%. Considering temporal and spatial effect, it is observed that the estimation of apartment price is more correct than the existing model. As the result of analyzing STAR model, the apartment price is affected as follows; area for apartment(-), years of apartment(-), dummy of low-rise(-), individual heating (-), city gas(-), dummy of reconstruction(+), stairs(+), size of complex(+). The results of other analysis method were the same. When estimating the price of real estate using STAR model, the government officials can improve policy efficiency and make reasonable investment based on the objective information by grasping trend of real estate market accurately.

• Communications for Statistical Applications and Methods
• /
• v.8 no.3
• /
• pp.777-786
• /
• 2001

The autoregressive models have been used to describe a wade variety of time series. Then the problem of determining the order in the times series model is very important in data analysis. We consider the Bayesian approach for finding the order of autoregressive(AR) error models using the latent variable which is motivated by Tanner and Wong(1987). The latent variables are combined with the coefficient parameters and the sequential steps are proposed to set up the prior of the latent variables. Markov chain Monte Carlo method(Gibbs sampler and Metropolis-Hasting algorithm) is used in order to overcome the difficulties of Bayesian computations. Three examples including AR(3) error model are presented to illustrate our proposed methodology.

• Communications for Statistical Applications and Methods
• /
• v.18 no.3
• /
• pp.319-331
• /
• 2011

In this paper, we propose automatic procedures for the model selection of various univariate time series data. Automatic model selection is important, especially in data mining with large number of time series, for example, the number (in thousands) of signals accessing a web server during a specific time period. Several methods have been proposed for automatic model selection of time series. However, most existing methods focus on linear time series models such as exponential smoothing and autoregressive integrated moving average(ARIMA) models. The key feature that distinguishes the proposed procedures from previous approaches is that the former can be used for both linear time series models and nonlinear time series models such as threshold autoregressive(TAR) models and autoregressive moving average-generalized autoregressive conditional heteroscedasticity(ARMA-GARCH) models. The proposed methods select a model from among the various models in the prediction error sense. We also provide an R package autots that implements the proposed automatic model selection procedures. In this paper, we illustrate these algorithms with the artificial and real data, and describe the implementation of the autots package for R.

• Communications for Statistical Applications and Methods
• /
• v.22 no.6
• /
• pp.557-573
• /
• 2015

In this paper we develop a Bayesian inference for a multiplicative double seasonal autoregressive (DSAR) model by implementing a fast, easy and accurate Gibbs sampling algorithm. We apply the Gibbs sampling to approximate empirically the marginal posterior distributions after showing that the conditional posterior distribution of the model parameters and the variance are multivariate normal and inverse gamma, respectively. The proposed Bayesian methodology is illustrated using simulated examples and real-world time series data.

• The Korean Journal of Applied Statistics
• /
• v.36 no.1
• /
• pp.21-32
• /
• 2023

The amount of money processed in medicine from the Korea Consumer Agency was studied by the various time series models. The medical data set from the Korea Consumer Agency were consisted of counseling, damage relief and conciliation. For the analysis of time series, autoregressive moving average model, vector autoregressive model and the transfer function model were used. We considered the stationarity and cross correlation function for the identification and fitting. As a result, the transfer function model showed a better prediction. Whereas, the vector autoregressive model also provided good information for the degree and duration of the influence of variables.

• Journal of the Korean Data and Information Science Society
• /
• v.20 no.5
• /
• pp.949-954
• /
• 2009

The autoregressive process is applied in this paper to kernel regression in order to infer nonlinear models for predicting responses. We propose a kernel method for the autoregressive data which estimates the mean function by kernel machines. We also present the model selection method which employs the cross validation techniques for choosing the hyper-parameters which affect the performance of kernel regression. Artificial and real examples are provided to indicate the usefulness of the proposed method for the estimation of mean function in the presence of autocorrelation between data.

• Journal of the Korean Data and Information Science Society
• /
• v.10 no.2
• /
• pp.397-404
• /
• 1999

For two-dimensional causal spatial autoregressive processes, we propose and illustrate a method for determining asymptotic simultaneous confidence regions using Yule-Walker, unbiased Yule-Walker and least squres estimators. The spectral density for first-order spatial autoregressive model are looked at in more detail. Finite sample properties based on simulation study we also presented.

• Communications for Statistical Applications and Methods
• /
• v.25 no.6
• /
• pp.659-671
• /
• 2018

In this study, we consider the extension of the heterogeneous autoregressive (HAR) model for realized volatility by incorporating a neural network (NN) structure. Since HAR is a linear model, we expect that adding a neural network term would explain the delicate nonlinearity of the realized volatility. Three neural network-based HAR models, namely HAR-NN, $HAR({\infty})-NN$, and HAR-AR(22)-NN are considered with performance measured by evaluating out-of-sample forecasting errors. The results of the study show that HAR-NN provides a slightly wider interval than traditional HAR as well as shows more peaks and valleys on the turning points. It implies that the HAR-NN model can capture sharper changes due to higher volatility than the traditional HAR model. The HAR-NN model for prediction interval is therefore recommended to account for higher volatility in the stock market. An empirical analysis on the multinational realized volatility of stock indexes shows that the HAR-NN that adds daily, weekly, and monthly volatility averages to the neural network model exhibits the best performance.