• Journal of KIISE:Databases
• /
• v.34 no.6
• /
• pp.483-492
• /
• 2007

Time series of conventional prediction techniques uses the model which is generated from the training step. This model is applied to new input data without any change. If this model is applied directly to stream data, the rate of prediction accuracy will be decreased. This paper proposes an stream data prediction technique using sliding window and regression. This technique considers the characteristic of time series which may be changed over time. It is composed of two steps. The first step executes a fractional process for applying input data to the regression model. The second step updates the model by using its information as new data. Additionally, the model is maintained by only recent data in a queue. This approach has the following two advantages. It maintains the minimum information of the model by using a matrix, so space complexity is reduced. Moreover, it prevents the increment of error rate by updating the model over time. Accuracy rate of the proposed method is measured by RME(Relative Mean Error) and RMSE(Root Mean Square Error). The results of stream data prediction experiment are performed by the proposed technique IMQR(Incremental Multiple Quadratic Regression) is more efficient than those of MLR(Multiple Linear Regression) and SVR(Support Vector Regression).

• Journal of Convergence for Information Technology
• /
• v.11 no.3
• /
• pp.7-13
• /
• 2021

This study develops an artificial intelligence prediction system for Fine particulate Matter(PM2.5) based on the deep learning algorithm GAN model. The experimental data are closely related to the changes in temperature, humidity, wind speed, and atmospheric pressure generated by the time series axis and the concentration of air pollutants such as SO2, CO, O3, NO2, and PM10. Due to the characteristics of the data, since the concentration at the current time is affected by the concentration at the previous time, a predictive model for recursive supervised learning was applied. For comparative analysis of the accuracy of the existing models, CNN and LSTM, the difference between observation value and prediction value was analyzed and visualized. As a result of performance analysis, it was confirmed that the proposed GAN improved to 15.8%, 10.9%, and 5.5% in the evaluation items RMSE, MAPE, and IOA compared to LSTM, respectively.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.7
• /
• pp.73-80
• /
• 2024

In this paper, we proposes a method to improve the accuracy of predicting the Korea Composite Stock Price Index (KOSPI) by combining topic modeling and Long Short-Term Memory (LSTM) neural networks. In this paper, we use the Latent Dirichlet Allocation (LDA) technique to extract ten major topics related to interest rate increases and decreases from financial news data. The extracted topics, along with historical KOSPI index data, are input into an LSTM model to predict the KOSPI index. The proposed model has the characteristic of predicting the KOSPI index by combining the time series prediction method by inputting the historical KOSPI index into the LSTM model and the topic modeling method by inputting news data. To verify the performance of the proposed model, this paper designs four models (LSTM_K model, LSTM_KNS model, LDA_K model, LDA_KNS model) based on the types of input data for the LSTM and presents the predictive performance of each model. The comparison of prediction performance results shows that the LSTM model (LDA_K model), which uses financial news topic data and historical KOSPI index data as inputs, recorded the lowest RMSE (Root Mean Square Error), demonstrating the best predictive performance.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.20 no.3
• /
• pp.59-73
• /
• 2021

There are many lives lost due traffic accidents, and which have not decreased despite advances in technology. In order to prevent traffic accidents, it is necessary to accurately forecast how they will change in the future. Until now, traffic accident-frequency forecasting has not been a major research field, but has been analyzed microscopically by traditional methods, mainly based on statistics over a previous period of time. Despite the recent introduction of AI to the traffic accident field, the focus is mainly on forecasting traffic flow. This study converts into time series data the records from 1,339,587 traffic accidents that occurred in Korea from 2014 to 2019, and uses the AI algorithm to forecast the frequency of traffic accidents based on driver's age and time of day. In addition, the forecast values and the actual values were compared and verified based on changes in the traffic environment due to COVID-19. In the future, these research results are expected to lead to improvements in policies that prevent traffic accidents.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2012.05a
• /
• pp.116-116
• /
• 2012

하천유역 내의 인자를 이용하여 댐의 하천유량(stream flow)을 예측하는 일은 수문특성의 연구와 자연재해에 대한 대비 및 수공구조물과 방재시설의 설계 시 중요한 역할을 한다. 이러한 연구는 과거부터 활발히 이루어졌으며, 아직도 보다 높은 정확도의 결과를 얻기 위해 많은 연구들이 이루어지고 있다. 특히 기존의 유역 내 자료를 통해 비선형적 모델인 인공신경망(artificial neural network)을 이용한 하천유량을 예측하는 연구 역시 활발히 이루어지고 있다. 본 연구의 목적은 여러 유역인자들 중 하천유량에 가장 영향을 미치는 변수를 추출하고 보다 정확한 예측모델을 구축하는 것이다. 기존의 입력자료 선정기법중의 하나인 상호정보량(mutual information)과 수문기상자료의 비선형 동역학적 성분을 추출하는 웨이블렛 변환(wavelet transform)을 사용하여 인공신경망에 적용시켰다. 인공신경망을 적용하는 경우, 수문자료에 있어서 변수의 선택과 자료의 상태가 강우예측의 결과에 큰 영향을 미친다. 이러한 변수의 선택에 있어서 상호정보량을 바탕으로 한 인공신경망 입력변수 선택기법이 많이 사용되고 있다. 일반적으로 시계열자료는 경향성(trend), 주기성(periodicity) 및 추계학적 성분(stochastic component)의 선형조합으로 가정될 수 있으며, 특히 경향성과 주기성은 시계열 모형을 위해 제거되어야 할 결정론적 성분으로 취급한다. 즉. 수문 기상자료에 포함되어 있는 경향성과 주기성과 같은 비선형 동역학적 잡음(nonlinear dynamical noise)을 제거하고 입력자료의 카오스적 거동을 보이는 성분을 분리하기 위해 웨이블렛 변환을 사용하였다. 대상유역은 한강 유역에 포함되어 있는 충주댐으로 선택하였다. 유역 내 다양한 인자들과 하천유량사이의 상호정보량을 구해 영향력이 가장 큰 변수를 추출하고, 그 자료를 웨이블렛 변환을 적용하여 인공신경망의 입력자료로 사용하였다. 본 논문에서는 위와 같은 과정을 이용해 추정한 하천유량 결과와 기존의 방법인 상호정보량을 이용해 인공신경망을 적용한 결과를 실제자료와 비교하였다.

• The Journal of the Korea Contents Association
• /
• v.19 no.1
• /
• pp.132-140
• /
• 2019

Recently, many cities around the world introduced and operated shared bicycle system to reduce the traffic and air pollution. Seoul also provides shared bicycle service called as "Ddareungi" since 2015. As the use of shared bicycle increases, the demand for bicycle in each station is also increasing. In addition to the restriction on budget, however, there are managerial issues due to the different demands of each station. Currently, while bicycle rebalancing is used to resolve the huge imbalance of demands among many stations, forecasting uncertain demand at the future is more important problem in practice. In this paper, we develop forecasting model for demand for Seoul shared bicycle using statistical time series analysis and apply our model to the real data. In particular, we apply Holt-Winters method which was used to forecast electricity demand, and perform sensitivity analysis on the parameters that affect on real demand forecasting.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.7
• /
• pp.29-37
• /
• 2023

Due to climate change, interest in crop production and distribution is increasing, and attempts are being made to use bigdata and AI to predict production volume and control shipments and distribution stages. Prediction of agricultural product imports not only affects prices, but also controls shipments of farms and distributions of distribution companies, so it is important information for establishing marketing strategies. In this paper, we create an artificial intelligence prediction model that predicts the future import volume based on the wholesale market melon import volume data disclosed by the agricultural statistics information system and evaluate its accuracy. We create prediction models using three models: the Neural Prophet technique, the Ensembled Neural Prophet model, and the GRU model. As a result of evaluating the performance of the model by comparing two major indicators, MAE and RMSE, the Ensembled Neural Prophet model predicted the most accurately, and the GRU model also showed similar performance to the ensemble model. The model developed in this study is published on the web and used in the field for 1 year and 6 months, and is used to predict melon production in the near future and to establish marketing and distribution strategies.

• Smart Media Journal
• /
• v.9 no.4
• /
• pp.26-35
• /
• 2020

Attempts to predict future stock prices have been studied steadily since the past. However, unlike general time-series data, financial time-series data has various obstacles to making predictions such as non-stationarity, long-term dependence, and non-linearity. In addition, variables of a wide range of data have limitations in the selection by humans, and the model should be able to automatically extract variables well. In this paper, we propose a 'sliding time step normalization' method that can normalize non-stationary data and LSTM autoencoder to compress variables from all variables. and 'moving transfer learning', which divides periods and performs transfer learning. In addition, the experiment shows that the performance is superior when using as many variables as possible through the neural network rather than using only 100 major financial variables and by using 'sliding time step normalization' to normalize the non-stationarity of data in all sections, it is shown to be effective in improving performance. 'moving transfer learning' shows that it is effective in improving the performance in long test intervals by evaluating the performance of the model and performing transfer learning in the test interval for each step.

• Journal of Advanced Navigation Technology
• /
• v.11 no.3
• /
• pp.252-258
• /
• 2007

There has been a lot of research done in scheme guaranteeing user's mobility and effective resources management to satisfy the requested by users in the wireless/mobile networks. In this paper, we propose a predictive resource allocation scheme based on ARMA(Auto Regressive Moving Average) prediction model to meet QoS requirements(handoff dropping rate) for guaranteeing users' mobility. The proposed scheme predicts the demanded amount of resource in the future time by ARMA time series prediction model, and then reserves it. The ARMA model can be used to take into account the correlation of future handoff resource demands with present and past handoff demands for provision of targeted handoff dropping rate. Simulation results show that the proposed scheme outperforms the existing RCS(Reserved channel scheme) in terms of handoff connection dropping rate and resource utilization.

• 한국해양학회지
• /
• v.27 no.1
• /
• pp.35-45
• /
• 1992

This paper is concerned with the development and application of a stochastic model for predicting E1 nino phenomenon. For this, first a general criterion for determining E1 nino phenomenon, including period and strength, which is based on partial sum of monthly sea surface temperatures (SST) anomalies, is proposed, Secondly, the annual fluctuations, periodicity and dependence of monthly mean of equatorial Pacific SST during the period 1951-1990 are analyzed. Based on these, time series nonlinear regression model for the prediction of SST have been derived. A statistical procedure for using the model to predict the SST have been derived. A statistical procedure for using the model to predict the SST level is also proposed.