• Industrial Engineering and Management Systems
• /
• 제7권1호
• /
• pp.9-22
• /
• 2008

As an extension of a previous work by the authors (Song and Esogbue, 2006), a new algorithm for automated modeling of nonstationary seasonal time series is presented in this paper. Issues relative to the methodology for building automatically seasonal time series models and periodic time series models are addressed. This is achieved by inspecting the trend, estimating the seasonality, determining the orders of the model, and estimating the parameters. As in our previous work, the major instruments used in the model identification process are correlograms of the modeling errors while the least square method is used for parameter estimation. We provide numerical illustrations of the performance of the new algorithms with respect to building both seasonal time series and periodic time series models. Additionally, we consider forecasting and exercise the models on some sample time series problems found in the literature as well as real life problems drawn from the retail industry. In each instance, the models are built automatically avoiding the necessity of any human intervention.

• Industrial Engineering and Management Systems
• /
• 제13권4호
• /
• pp.357-368
• /
• 2014

Since its birth in 1993, fuzzy time series have seen different classes of models designed and applied, such as fuzzy logic relation and rule-based models. These models have both advantages and disadvantages. The major drawbacks with these two classes of models are the difficulties encountered in identification and analysis of the model. Therefore, there is a strong need to explore new alternatives and this is the objective of this paper. By transforming a fuzzy number to a real number via integrating the inverse of the membership function, new autoregressive models can be developed to fit the observation values of a fuzzy time series. With the new models, the issues of model identification and parameter estimation can be addressed; and trends, seasonalities and multivariate fuzzy time series could also be modeled with ease. In addition, asymptotic behaviors of fuzzy time series can be inspected by means of characteristic equations.

• Communications for Statistical Applications and Methods
• /
• 제18권3호
• /
• 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.

• Journal of the Korean Statistical Society
• /
• 제30권3호
• /
• pp.387-406
• /
• 2001

Adaptive estimators for a class of nonlinear time series models has been proposed by several authors. Koul and Schick(1997) proposed the adaptive estimators without sample splitting for location-type time series models. They also showed by simulation that the adaptive estimators without sample splitting have smaller mean squared errors than those of the adaptive estimators with sample splitting. the present paper generalized the result in a case of location-scale type nonlinear time series models by simulation.

• 응용통계연구
• /
• 제20권2호
• /
• pp.219-227
• /
• 2007

시계열 모형은 통신망 트래픽의 예측과 분석에 유용하게 쓰여 왔다. 본 논문에서는 통신망 트래픽의 예측을 위하여 다양한 시계열 모형을 소개하고 성능평가를 하고자 한다. 이를 위하여 실제 통신망 트래픽 자료에 선형 및 비선형 시계열모형을 적합 시키고 비선형 시계열모형이 선형 시계열 모형보다 예측의 정확도가 우수함을 보이고자 한다.

• 한국수자원학회논문집
• /
• 제35권5호
• /
• pp.631-638
• /
• 2002

시계열의 예측은 통상 추계학적 모형에 의해 수행하여 왔다. 그러나 본 연구에서는 퍼지 개념을 이용한 퍼지 시계열 모형에 의해 강수량 예측을 수행하였다. 기존에 제안된 퍼지 시계열 모형을 이용하여 예측을 수행하고, 예측 능력을 향상시키기 위하여 퍼지 시계열과 뉴로-퍼지 시스템을 연계한 새로운 방법론을 제안하여 상호 비교ㆍ분석하였다. 이를 위하여 미국 일리노이주의 강수량 시계열 예측에 적용하였으며, 예측 결과, 기존의 모형보다 본 연구에서 제안한 방법론의 결과가 더 정확함을 알 수 있었다.

• 시큐리티연구
• /
• 제30호
• /
• pp.139-169
• /
• 2012

본 연구는 살인, 강도, 강간, 절도, 폭력 등 주요 범죄를 예측할 수 있는 시계열 모형을 도출하고 이를 이용한 주요 범죄의 발생 전망을 파악하여 범죄 발생에 대한 과학적인 치안정책 수립에 기여하는데 그 목적이 있다. 이와 같은 목적을 달성하기 위하여 2002년부터 2010년까지의 살인, 강도, 강간, 절도, 폭력 등 주요범죄에 대한 월별 발생건수를 IBM PASW(SPSS) 19.0을 사용하여 주요 범죄의 시계열 예측모형을 규명하기 위한 시계열 모형생성(C), 주요 범죄의 시계열 예측모형에 대한 정확도 규명을 위한 시계열 모형생성(C) 및 시계열 순차도표(N)를 실시하였다. 이와 같은 연구목적과 연구방법을 통하여 도출한 연구결과는 다음과 같다. 첫째, 살인, 강도, 강간, 절도, 폭력 등 주요 범죄에 대한 시계열 예측모형은 각각 단순계절, Winters 승법, ARIMA(0,1,1)(0,1,1), ARIMA(1,1,0)(0,1,1), 단순계절로 나타났다. 둘째, 살인, 강도, 강간, 절도, 폭력 등 주요 범죄에 대하여 시계열 예측모형을 이용한 주요 범죄에 대한 단기적 발생 전망이 가능한 것으로 나타났다. 이러한 연구결과를 토대로 범죄 발생에 대한 지속적인 시계열 예측모형 제시, 분기별, 연도별 범죄 발생건수를 기초로 하는 중 장기 시계열 예측모형에 대한 관심이 요구된다.

• 응용통계연구
• /
• 제29권1호
• /
• pp.1-12
• /
• 2016

금융시계열은 일반 시계열과는 차별적으로 stylized facts로 불리는 특징을 가지고 있다. 이 특징들은 급첨 성질, 비정규분포, 변동성 집중 및 비대칭성을 포함한다. 이러한 특징들을 설명하기 위해서는 기존의 선형 ARMA 모형에서 벗어난 특수한 모형이 필요하게 되었다. 본 논문은 변동성 모형인 GARCH 형태의 모형을 중심으로 특수 금융시계열 모형들을 소개하고 연관된 통계적 이슈들에 대해 가능한 최근 연구를 중심으로 폭 넓게 조망하고 있다.

• Communications for Statistical Applications and Methods
• /
• 제28권1호
• /
• pp.39-57
• /
• 2021

Time series prediction is an area of great interest to many people. Algorithms for time series prediction are widely used in many fields such as stock price, temperature, energy and weather forecast; in addtion, classical models as well as recurrent neural networks (RNNs) have been actively developed. After introducing the attention mechanism to neural network models, many new models with improved performance have been developed; in addition, models using attention twice have also recently been proposed, resulting in further performance improvements. In this paper, we consider time series prediction by introducing attention twice to an RNN model. The proposed model is a method that introduces H-attention and T-attention for output value and time step information to select useful information. We conduct experiments on stock price, temperature and energy data and confirm that the proposed model outperforms existing models.

• 무역학회지
• /
• 제46권6호
• /
• pp.191-209
• /
• 2021

In this paper, various time series analysis models and machine learning models are presented for long-term prediction of export growth rate, and the prediction performance is compared and reviewed by RMSE and MAE. Export growth rate is one of the major economic indicators to evaluate the economic status. And It is also used to predict economic forecast. The export growth rate may have a negative (-) value as well as a positive (+) value. Therefore, Instead of using the ReLU function, which is often used for time series prediction of deep learning models, the PReLU function, which can have a negative (-) value as an output value, was used as the activation function of deep learning models. The time series prediction performance of each model for three types of data was compared and reviewed. The forecast data of long-term prediction of export growth rate was deduced by three forecast methods such as a fixed forecast method, a recursive forecast method and a rolling forecast method. As a result of the forecast, the traditional time series analysis model, ARDL, showed excellent performance, but as the time period of learning data increases, the performance of machine learning models including LSTM was relatively improved.