Kyong Joo Oh

Using rough set to support investment strategies of real-time trading in futures market

Finding proper investment strategies in futures market has been a hot issue to everyone involved in major financial markets around the world. However, it is a very difficult problem because of intrinsic unpredictability of the market. What makes things more complicated is the advent of real-time trading due to recent striking advancement of electronic communication technology. The real-time data imposes many difficult tasks to futures market analyst since it provides too much information to be analyzed for an instant. Thus it is inevitable for an analyst to resort to a rule-based trading system for making profits, which is usually done by the help of diverse technical indicators. In this study, we propose using rough set to develop an efficient real-time rule-based trading system (RRTS). In fact, we propose a procedure for building RRTS which is based on rough set analysis of technical indicators. We examine its profitability through an empirical study.

Facilitating cross-selling in a mobile telecom market to develop customer classification model based on hybrid data mining techniques

Graphical abstractDisplay Omitted Research highlights? In this study, we propose a customer classification model. ? The model may be used for facilitating cross-selling in a mobile telecom market. ? The various data mining techniques are applied to our proposed model in two steps. ? In the first step, our model predicts the purchase of new products. ? In the second step, our model compromises prediction probabilities by using GA. As the competition between mobile telecom operators becomes severe, it becomes critical for operators to diversify their business areas. Especially, the mobile operators are turning from traditional voice communication to mobile value-added services (VAS), which are new services to generate more average revenue per user (ARPU). That is, cross-selling is critical for mobile telecom operators to expand their revenues and profits. In this study, we propose a customer classification model, which may be used for facilitating cross-selling in a mobile telecom market. Our model uses the cumulated data on the existing customers including their demographic data and the patterns for using old products or services to find new products and services with high sales potential. The various data mining techniques are applied to our proposed model in two steps. In the first step, several classification techniques such as logistic regression, artificial neural networks, and decision trees are applied independently to predict the purchase of new products, and each model produces the results of their prediction as a form of probabilities. In the second step, our model compromises all these probabilities by using genetic algorithm (GA), and makes the final decision for a target customer whether he or she would purchase a new product. To validate the usefulness of our model, we applied it to a real-world mobile telecom companys case in Korea. As a result, we found that our model produced high-quality information for cross-selling, and that GA in the second step contributed to significantly improve the performance.

Using ridge regression with genetic algorithm to enhance real estate appraisal forecasting

This study considers real estate appraisal forecasting problem. While there is a great deal of literature about use of artificial intelligence and multiple linear regression for the problem, there has been always controversy about which one performs better. Noting that this controversy is due to difficulty finding proper predictor variables in real estate appraisal, we propose a modified version of ridge regression, i.e., ridge regression coupled with genetic algorithm (GA-Ridge). In order to examine the performance of the proposed method, experimental study is done for Korean real estate market, which verifies that GA-Ridge is effective in forecasting real estate appraisal. This study addresses two critical issues regarding the use of ridge regression, i.e., when to use it and how to improve it.

An early warning system for detection of financial crisis using financial market volatility

Abstract: This study proposes an early warning system (EWS) for detection of financial crisis with a daily financial condition indicator (DFCI) designed to monitor the financial markets and provide warning signals. The proposed EWS differs from other commonly used EWSs in two aspects: (i) it is based on dynamic daily movements of the financial markets; and (ii) it is established as a pattern classifier, which identifies predefined unstable states in terms of financial market volatility. Indeed it issues warning signals on a daily basis by judging whether the financial market has entered a predefined unstable state or not. The major strength of a DFCI is that it can issue timely warning signals while other conventional EWSs must wait for the next round input of monthly or quarterly information. Construction of a DFCI consists of two steps where machine learning algorithms are expected to play a significant role, i.e. (i) establishing sub-DFCIs on various daily financial variables by an artificial neural network, and (ii) integrating the sub-DFCIs into an integrated DFCI by a genetic algorithm. The DFCI for the Korean financial market is built as an empirical case study.

An early warning system for global institutional investors at emerging stock markets based on machine learning forecasting

At local emerging stock markets such as Korea, Hong Kong, Singapore and Taiwan, global institutional investors (GII) comprised of global mutual funds, offshore funds, and hedge funds play a key role and more often than not cause severe turmoil via massive selling. Thus, for the concerned local governments or private and institutional investors, it is quite necessary to monitor the behavior of GII against a sudden pullout. The main aim of this article is to propose an early warning system (EWS) which purposes issuing warning signal against the possible massive selling of GII at the local market. For this, we introduce machine learning algorithm which forecasts the behavior of GII by predicting future conditions. Technically, this EWS is an advanced form of the EWS developed by Oh et al. [Oh, K. J., Kim, T. Y., & Kim, C. (2006). An early warning system for detection of financial crisis using financial market volatility. Expert Systems, 23, 83-98] which issues a warning based on classifying present conditions. This study is empirically done for the Korean stock market.

Usefulness of support vector machine to develop an early warning system for financial crisis

Oh, Kim, and Kim (2006a), Oh, Kim, Kim, and Lee (2006b) proposed a classification approach for building an early warning system (EWS) against potential financial crises. This EWS classification approach has been developed mainly for monitoring daily financial market against its abnormal movement and is based on the newly-developed crisis hypothesis that financial crisis is often self-fulfilling because of herding behavior of the investors. This article extends the EWS classification approach to the traditional-type crisis, i.e., the financial crisis is an outcome of the long-term deterioration of the economic fundamentals. It is shown that support vector machine (SVM) is an efficient classifier in such case.

An early warning system for financial crisis using a stock market instability index

: This paper proposes to utilize a stock market instability index (SMII) to develop an early warning system for financial crisis. The system focuses on measuring the differences between the current market conditions and the conditions of the past when the market was stable. Technically the system evaluates the current time series against the past stable time series modelled by an asymptotic stationary autoregressive model via artificial neural networks. Advantageously accessible to extensive resources, the system turns out better results than the conventional system which detects similarities between the conditions of the current market and the conditions of previous markets that were in crisis. Therefore, it should be considered as a more advanced tool to prevent financial crises than the conventional one. As an empirical example, an SMII for the Korean stock market is developed in order to demonstrate its potential usefulness as an early warning system.

Using neural networks to support early warning system for financial crisis forecasting

This study deals with the construction process of a daily financial condition indicator (DFCI), which can be used as an early warning signal using neural networks and nonlinear programming. One of the characteristics in the proposed indicator is to establish an alarm zone in the DFCI, which plays a role of predicting a potential financial crisis. The previous financial condition indicators based on statistical methods are developed such that they examine whether a crisis will be break out within 24 months. In this study, however, the alarm zone makes it possible for the DFCI to forecast an unexpected crisis on a daily basis and then issue an early warning signal. Therefore, DFCI involves daily monitoring of the evolution of the stock price index, foreign exchange rate and interest rate, which tend to exhibit unusual behaviors preceding a possible crisis. Using nonlinear programming, the procedure of DFCI construction is completed by integrating three sub-DFCIs, based on each financial variable, into the final DFCI. The DFCI for Korean financial market will be established as an empirical study. This study then examines the predictability of alarm zone for the financial crisis forecasting in Korea.

Using decision tree to develop a soil ecological quality assessment system for planning sustainable construction

Research highlights? This study proposes a soil ecological quality assessment system. ? The system uses forward and backward DT models under GIS-based spatial analysis. ? The system may examine conservation and development areas strictly. ? The prediction results can be used for planning mega-construction projects. Soil ecology is the foundation of the entire biosphere and plays a significant role in global ecosystems. Soil ecology is important in the decision-making aspects of mega-construction projects. Despite its significance, soil ecological quality is not normally included in environmental impact assessments for sustainable development. This study develops and presents a new expert system to assess soil microbial diversity as an indicator of soil ecology quality using decision tree (DT) algorithms and GIS (geographic information system)-based spatial analysis. Our modeling results show that forward and backward DT models provide development-oriented and conservation-oriented information maps. To resolve potential conflicts by the different model predictions, a new mapping approach was developed for identifying strict conservation and potential development areas. These results suggest that the newly developed soil ecological quality assessment system can be used for planning mega-construction projects.

Developing time-based clustering neural networks to use change-point detection: Application to financial time series

This study suggests time-based clustering models integrating change-point detection and neural networks, and applies them to financial time series forecasting. The basic concept of the proposed models is to obtain intervals divided by change points, to identify them as change-point groups, and to involve them in the forecasting model. The proposed models consist of two stages. The first stage, the clustering neural network modeling stage, is to detect successive change points in the dataset, and to forecast change-point groups with backpropagation neural networks (BPNs). In this stage, three change-point detection methods are applied and compared. They are: (1) the parametric approach, (2) the nonparametric approach, and (3) the model-based approach. The next stage is to forecast the final output with BPNs. Through the application to financial time series forecasting, we compare the proposed models with a neural network model alone and, in addition, determine which of three change-point detection methods performs better. Furthermore, we evaluate whether the proposed models play a role in clustering to reflect the time. Finally, this study examines the predictability of the integrated neural network models based on change-point detection.