• Structural Engineering and Mechanics
• /
• 제45권5호
• /
• pp.693-702
• /
• 2013

Recently, artificial intelligence tools are most used for structural engineering and mechanics. In order to predict reserve prices and prices of awards, this study proposed a novel regression prediction model by the intelligent Kalman filtering method. An artificial intelligent multiple regression model was established using categorized data and then a prediction model using intelligent Kalman filtering. The rather precise construction bid price model was selected for the purpose of increasing the probability to win bids in the simulation example.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.524-524
• /
• 2000

We have been developing a method to build one-step-ahead prediction models for time series using genetic programming (GP). Our model building method consists of two stages. In the first stage, functional forms of the models are inherited from their parent models through crossover operation of GP. In the second stage, the parameters of the newborn model arc optimized based on an iterative method just like the back propagation. The proposed method has been applied to various kinds of time series problems. An application to the seismic ground motion was presented in the KACC'99, and since then the method has been improved in many aspects, for example, additions of new node functions, improvements of the node functions, and new exploitations of many kinds of mutation operators. The new ideas and trials enhance the ability to generate effective and complicated models and reduce CPU time. Today, we will present a couple of financial applications, espc:cially focusing on gold price prediction in Tokyo market.

• 지능정보연구
• /
• 제29권2호
• /
• pp.129-148
• /
• 2023

블록체인 기술이 적용되어 있는 암호화폐는 높은 가격 변동성을 가지며 투자자 및 일반 대중으로부터 큰 관심을 받아왔다. 이러한 관심을 바탕으로 암호화폐를 비롯한 투자상품의 미래가치를 예측하기 위한 연구가 이루어지고 있으나 예측모형에 대한 설명력 및 해석 가능성이 낮아 실무에서 활용하기 어렵다는 비판을 받아왔다. 본 연구에서는 암호화폐 가격 예측모형의 성과를 향상시키기 위해 금융투자상품의 가치평가에 활용되는 기술적 지표들과 함께 투자자의 사회적 관심도를 반영할 수 있는 구글 키워드 검색량 데이터를 사용하고 설명 가능한 인공지능을 적용하여 모형에 대한 해석을 제공하고자 한다. 최근 금융 시계열 분야에서 예측성과의 우수성을 인정받고 있는 LSTM(Long Short Term Memory)과 CNN(Convolutional Neural Networks)을 활용하고, 'bitcoin'을 검색어로 하는 구글 검색량 데이터를 적용해 일주일 후의 가격 등락 예측모형을 구축하였다. LSTM과 CNN을 활용해 구축한 모형들이 높은 예측성능을 보였으며 구글 검색량을 반영한 모형에서 더 높은 예측성과를 확인할 수 있었다. 딥러닝 모형의 해석 가능성 및 설명력을 위해 XAI의 SHAP 기법을 적용한 결과, 구글 검색량과 함께 과매수, 과매도 정도를 파악할 수 있는 지표들이 모형의 의사결정에 가장 큰 영향들을 미치고 있음을 파악할 수 있었다. 본 연구는 암호화폐 가격 등락 예측에 있어 전통적으로 시계열 예측에 우수한 성과를 인정받고 있는 LSTM뿐만 아니라 이미지 분류에서 높은 예측성과를 보이는 딥러닝 기법인 CNN 또한 우수한 예측성능을 보일 수 있음을 확인하였으며, XAI를 통해 예측모형에 대한 해석을 제공하고, 대중의 심리를 반영하는 정보 중 하나인 구글 검색량을 활용해 예측성과를 향상시킬 수 있다는 것을 확인했다는 점에서 의의가 있다.

• 한국시뮬레이션학회논문지
• /
• 제17권3호
• /
• pp.19-26
• /
• 2008

본 논문에서는 중기 국채(Treasure Note; T-Note)의 실제 자료를 이용하여 채권 가격에 대한 이자율을 예측하는 동적인 예측 알고리즘을 제안하고 있다. 제안한 알고리즘은 이자율 기간 구조를 근본으로 하고 있으며 표준 위너 과정(standard Wiener process)과 같은 다양한 금융 모형의 대안으로 활용 가능하다. 본 논문에서는 실제 자료의 누적 분포 함수(Cumulative Distribution Function; CDF)를 이용하여 이자율을 측정하였으며 CDF는 수치적 방법인 보간법 중에 자주 활용되는 내츄럴 큐빅 스플라인(natural cubic spline; NCS)방법을 통하여 얻었다. 위에서 얻은 CDF를 통하여 난수 생성기법(random number generation scheme; RNGS)을 이용하여 채권의 가격를 계산하였다. 컴퓨터 시뮬레이션을 통해 얻은 실험결과로부터 제안된 예측 알고리즘에서 엄밀도(precision)의 낮은 값을 얻음으로써 채권의 가치가 더욱 예리하고 정확하게 평가되었음을 확인할 수 있었으며, 이는 매우 근거 있는 예측이라 할 수 있다.

• 대한지리학회지
• /
• 제48권2호
• /
• pp.272-287
• /
• 2013

본 연구는 부동산 가격 추정을 위해 자주 활용되는 헤도닉 가격모형(Hedonic Pricing Model)에서의 설명변수, 즉 지가형성요인의 선별 중요성에 대해 기술하고, 이러한 지가형성요인 및 그 효과가 시간의 경과에 따라 어떻게 변화하는지 실증적으로 검토하였다. 전주시를 사례 지역으로 하여 17년간 반복 측정된 표준지 공시지가를 분석 대상으로 하였으며, 자료가 가지는 포섭구조(nested structure) 및 종단성(longitudinal characteristics)을 고려하여 3개 수준으로 구성된 다수준모형(multi-level model)을 설정하여 적합 정도를 평가하였다. 지가형성요인은 공시지가 산정시 활용되는 헤도닉 가격모형의 일종인 비준표(比準表)에 포함된 항목을 중심으로 살펴보았다. 분석 결과, 17년간의 지가 변동 추세는 전주시 세부지역별로 상승 또는 하락하는 등 지역마다 다른 추이를 보였으며, 따라서 종단효과의 모델 반영이 필요함이 확인되었다. 또한 일반적으로 중요하다고 여겨지는 지가형성요인 중 유의하지 않은 요인이 발견되었으며, 특정 시점에서 영향력이 상당히 큰 것으로 판명된 지가형성요인도 시간의 경과에 따라 그 영향력이 약해지는가 하면, 반대로 지가에 미치는 영향력이 초기에는 미약하였으나 점차 뚜렷해지는 요인이 파악되었다. 향후 헤도닉 모형 적용시 이러한 지가형성요인의 동태성을 모델의 구성요소로 고려할 경우 보다 정확한 가격 추정이 가능해질 것이다.

• 한국정보시스템학회지:정보시스템연구
• /
• 제30권2호
• /
• pp.147-163
• /
• 2021

Purpose The study aims to predict real estate prices by utilizing regional characteristics. Since real estate has the characteristic of immobility, the characteristics of a region have a great influence on the price of real estate. In addition, real estate prices are closely related to economic development and are a major concern for policy makers and investors. Accurate house price forecasting is necessary to prepare for the impact of house price fluctuations. To improve the performance of our predictive models, we applied LSTM, a widely used deep learning technique for predicting time series data. Design/methodology/approach This study used time series data on real estate prices provided by the Ministry of Land, Infrastructure and Transport. For time series data preprocessing, HP filters were applied to decompose trends and SOM was used to cluster regions with similar price directions. To build a real estate price prediction model, SVR and LSTM were applied, and the prices of regions classified into similar clusters by SOM were used as input variables. Findings The clustering results showed that the region of the same cluster was geographically close, and it was possible to confirm the characteristics of being classified as the same cluster even if there was a price level and a similar industry group. As a result of predicting real estate prices in 1, 2, and 3 months, LSTM showed better predictive performance than SVR, and LSTM showed better predictive performance in long-term forecasting 3 months later than in 1-month short-term forecasting.

• 농촌계획
• /
• 제30권2호
• /
• pp.1-11
• /
• 2024

Farmland transaction prices are affected by various factors such as politics, society, and the economy. The purpose of this study is to identify multiple factors that affect the farmland transaction price due to changes in the actual transaction price of farmland by farmland unit from 2016 to 2020. There are several previous studies analyzed the determinants of farmland transaction prices by considering spatial dependency. However, in the case of land transactions where the time and space of the transaction affect simultaneously, if only spatial dependence is considered, there is a limitation in that it cannot reflect spatial dependence that occurs over time. In order to solve these limitations, To address these limitations, this study builds a spatio-temporal autoregressive model that simultaneously considers spatial and temporal dependencies using farmland transactions in Jinju City as an example. As a result of the analysis, it was confirmed that there was significant spatio-temporal dependence in farmland transactions within the previous 30 days. This means that if the previous farmland transaction was carried out at a high price, it has a spatio-temporal spillover effect that indirectly affects the increase in the price of other nearby farmland transactions. The study also found that various location attributes and socioeconomic attributes have a significant impact on farmland transaction prices. The spatio-temporal autoregressive model of farmland prices constructed in this study can be used to improve the prediction accuracy of farmland prices in the farmland transaction market in the future, and it is expected to be useful in drawing policy implications for stabilizing farmland prices

• 유통과학연구
• /
• 제22권4호
• /
• pp.11-22
• /
• 2024

Purpose: This research delves into the various factors that influence the performance of restaurant businesses on social commerce platforms in Bangkok, Thailand. The study considers both internal and external factors, including but not limited to business characteristics and location. Moreover, this research also analyzes the effects of employing multiple social commerce platforms on business efficiency and explores the underlying reasons for such effects. Research design, data, and methodology: Restaurants can be classified into different price ranges: low, medium, and high. To further investigate, we employed natural language processing AI to analyze online reviews and evaluate algorithm performance using machine learning techniques. We aimed to develop a model to gauge customer satisfaction with restaurants across different price categories effectively. Results: According to the research findings, several factors significantly impact restaurant groups in the low and mid-price ranges. Among these factors are population density and the number of seats at the restaurant. On the other hand, in the mid-and high-price ranges, the price levels of the food and drinks offered by the restaurant play a crucial role in determining customer satisfaction. Furthermore, the correlation between different social commerce platforms can significantly affect the business performance of high-price range restaurant groups. Finally, the level of online review sentiment has been found to influence customer decision-making across all restaurant types significantly. Conclusions: The study emphasizes that restaurants' characteristics based on their price level differ significantly, and social commerce platforms have the potential to affect one another. It is worth noting that the sentiment expressed in online reviews has a more significant impact on customer decision-making than any other factor, regardless of the type of restaurant in question.

• The Journal of Asian Finance, Economics and Business
• /
• 제6권3호
• /
• pp.19-26
• /
• 2019

The paper aims to measure stock price volatility on Ho Chi Minh stock exchange (HSX). We apply symmetric models (GARCH, GARCH-M) and asymmetry (EGARCH and TGARCH) to measure stock price volatility on HSX. We used time series data including the daily closed price of VN-Index during 1/03/2001-1/03/2019 with 4375 observations. The results show that GARCH (1,1) and EGARCH (1,1) models are the most suitable models to measure both symmetry and asymmetry volatility level of VN-Index. The study also provides evidence for the existence of asymmetric effects (leverage) through the parameters of TGARCH model (1,1), showing that positive shocks have a significant effect on the conditional variance (volatility). This result implies that the volatility of stock returns has a big impact on future market movements under the impact of shocks, while asymmetric volatility increase market risk, thus increase the attractiveness of the stock market. The research results are useful reference information to help investors in forecasting the expected profit rate of the HSX, and also the risks along with market fluctuations in order to take appropriate adjust to the portfolios. From this study's results, we can see risk prediction models such as GARCH can be better used in risk forecasting especially.

• 한국정보시스템학회지:정보시스템연구
• /
• 제27권3호
• /
• pp.123-140
• /
• 2018

Purpose The purpose of this study is to explore the optimal trading frequency which is useful for stock price prediction by using deep learning for charting image data. We also want to identify the appropriate time for accurate forecasting of stock price when performing pattern analysis. Design/methodology/approach In order to find the optimal trading frequency patterns and forecast timings, this study is performed as follows. First, stock price data is collected using OpenAPI provided by Daishin Securities, and candle chart images are created by data frequency and forecasting time. Second, the patterns are generated by the charting images and the learning is performed using the CNN. Finally, we find the optimal trading frequency patterns and forecasting timings. Findings According to the experiment results, this study confirmed that when the 10 minute frequency data is judged to be a decline pattern at previous 1 tick, the accuracy of predicting the market frequency pattern at which the market decreasing is 76%, which is determined by the optimal frequency pattern. In addition, we confirmed that forecasting of the sales frequency pattern at previous 1 tick shows higher accuracy than previous 2 tick and 3 tick.