Jun 2009, Volume 3 Issue 2

  • Select all
    Shuo BAI, Shouyang WANG, Lean YU, Aoying ZHOU
    Di WU, Gabriel Pui Cheong FUNG, Jeffrey Xu YU, Qi PAN

    There are many real applications existing where the decision making process depends on a model that is built by collecting information from different data sources. Let us take the stock market as an example. The decision making process depends on a model which that is influenced by factors such as stock prices, exchange volumes, market indices (e.g. Dow Jones Index), news articles, and government announcements (e.g., the increase of stamp duty). Yet Nevertheless, modeling the stock market is a challenging task because (1) the process related to market states (rise state/drop state) is a stochastic process, which is hard to capture using the deterministic approach, and (2) the market state is invisible but will be influenced by the visible market information, like stock prices and news articles. In this paper, we propose an approach to model the stock market process by using a Non-homogeneous Hidden Markov Model (NHMM). It takes both stock prices and news articles into consideration when it is being computed. A unique feature of our approach is event driven. We identify associated events for a specific stock using a set of bursty features (keywords), which has a significant impact on the stock price changes when building the NHMM. We apply the model to predict the trend of future stock prices and the encouraging results indicate our proposed approach is practically sound and highly effective.

    Serge HAYWARD

    Considering the effect of economic agents’ preferences on their actions, the relationships between conventional summary statistics and forecast profits are investigated. An analytical examination of loss function families demonstrates that investors’ utility maximisation is determined by their risk attitudes. In computational settings, stock traders’ fitness is assessed in response to a slow step increase in the value of the risk aversion coefficient. The experiment rejects the claims that the accuracy of the forecast does not depend upon which error-criteria are used and that none of them is related to the profitability of the forecast. The profitability of networks trained with L6 loss function appeared to be statistically significant and stable, although links between the loss functions and the accuracy of forecasts were less conclusive.

    Lean YU, Shouyang WANG, Kin Keung LAI

    The slow convergence of back-propagation neural network (BPNN) has become a challenge in data-mining and knowledge discovery applications due to the drawbacks of the gradient descent (GD) optimization method, which is widely adopted in BPNN learning. To solve this problem, some standard optimization techniques such as conjugategradient and Newton method have been proposed to improve the convergence rate of BP learning algorithm. This paper presents a heuristic method that adds an adaptive smoothing momentum term to original BP learning algorithm to speedup the convergence. In this improved BP learning algorithm, adaptive smoothing technique is used to adjust the momentums of weight updating formula automatically in terms of “3 σ limits theory.” Using the adaptive smoothing momentum terms, the improved BP learning algorithm can make the network training and convergence process faster, and the network’s generalization performance stronger than the standard BP learning algorithm can do. In order to verify the effectiveness of the proposed BP learning algorithm, three typical foreign exchange rates, British pound (GBP), Euro (EUR), and Japanese yen (JPY), are chosen as the forecasting targets for illustration purpose. Experimental results from homogeneous algorithm comparisons reveal that the proposed BP learning algorithm outperforms the other comparable BP algorithms in performance and convergence rate. Furthermore, empirical results from heterogeneous model comparisons also show the effectiveness of the proposed BP learning algorithm.

    Heping PAN, Imad HAIDAR, Siddhivinayak KULKARNI

    This paper documents a systematic investigation on the predictability of short-term trends of crude oil prices on a daily basis. In stark contrast with longer-term predictions of crude oil prices, short-term prediction with time horizons of 1-3 days posits an important problem that is quite different from what has been studied in the literature. The problem of such short-term predicability is tackled through two aspects. The first is to examine the existence of linear or nonlinear dynamic processes in crude oil prices. This sub-problem is addressed with statistical analysis involving the Brock-Dechert-Scheinkman test for nonlinearity. The second aspect is to test the capability of artificial neural networks (ANN) for modeling the implicit nonlinearity for prediction. Four experimental models are designed and tested with historical data: (1) using only the lagged returns of filtered crude oil prices as input to predict the returns of the next days; this is used as the benchmark, (2) using only the information set of filtered crude oil futures price as input, (3) combining the inputs from the benchmark and second models, and (4) combing the inputs from the benchmark model and the intermarket information. In order to filter out the noise in the original price data, the moving averages of prices are used for all the experiments. The results provided sufficient evidence to the predictability of crude oil prices using ANN with an out-of-sample hit rate of 80%, 70%, and 61% for each of the next three days’ trends.

    Dongling ZHANG, Yong SHI, Yingjie TIAN, Meihong ZHU

    An extensive review for the recent developments of multiple criteria linear programming data mining models is provided in this paper. These researches, which include classification and regression methods, are introduced in a systematic way. Some applications of these methods to real-world problems are also involved in this paper. This paper is a summary and reference of multiple criteria linear programming methods that might be helpful for researchers and applications in data mining.

    Ligang ZHOU, Kin Keung LAI

    In practice, there are many binary classification problems, such as credit risk assessment, medical testing for determining if a patient has a certain disease or not, etc. However, different problems have different characteristics that may lead to different difficulties of the problem. One important characteristic is the degree of imbalance of two classes in data sets. For data sets with different degrees of imbalance, are the commonly used binary classification methods still feasible? In this study, various binary classification models, including traditional statistical methods and newly emerged methods from artificial intelligence, such as linear regression, discriminant analysis, decision tree, neural network, support vector machines, etc., are reviewed, and their performance in terms of the measure of classification accuracy and area under Receiver Operating Characteristic (ROC) curve are tested and compared on fourteen data sets with different imbalance degrees. The results help to select the appropriate methods for problems with different degrees of imbalance.

    Dabin ZHANG, Lean YU, Shouyang WANG, Yingwen SONG

    A new clustering analysis method based on the pseudo parallel genetic algorithm (PPGA) is proposed for business cycle indicator selection. In the proposed method, the category of each indicator is coded by real numbers, and some illegal chromosomes are repaired by the identification and restoration of empty class. Two mutation operators, namely the discrete random mutation operator and the optimal direction mutation operator, are designed to balance the local convergence speed and the global convergence performance, which are then combined with migration strategy and insertion strategy. For the purpose of verification and illustration, the proposed method is compared with the K-means clustering algorithm and the standard genetic algorithms via a numerical simulation experiment. The experimental result shows the feasibility and effectiveness of the new PPGA-based clustering analysis algorithm. Meanwhile, the proposed clustering analysis algorithm is also applied to select the business cycle indicators to examine the status of the macro economy. Empirical results demonstrate that the proposed method can effectively and correctly select some leading indicators, coincident indicators, and lagging indicators to reflect the business cycle, which is extremely operational for some macro economy administrative managers and business decision-makers.

    Guihuan ZHENG, Xun ZHANG, Wei SHANG, Shanying XU

    In this paper, a financial early warning information system is developed based on the multi-dimensional climate approach that is featured with a multi-dimensional index construction and the relevant multi-dimensional analysis. Requirement analysis and design issues of building an information system supporting this multi-dimensional climate approach are discussed in detail. And a case using this system to study the macro financial issues is presented to illustrate how the proposed multi-dimensional approach works in the information system we design. This research is an interdisciplinary work of economic theories, macro financial empirical studies, and software engineering. With advanced macro financial early warning theories implemented in a web application, the Macro Financial Early Warning System (FEWS) developed in this research has been proved to be effective in a trial running in the Forecasting research institute of the Chinese Academy of Sciences.

    Feng WANG, Keren DONG, Xiaotie DENG

    This paper provides an overview of research and development in algorithmic trading and discusses key issues involved in the current effort on its improvement, which would be of great value to traders and investors. Some current systems for algorithmic trading are introduced, together with some illustrations of their functionalities. We then present our platform named FiSimn and discuss its overall design as well as some experimental results in user strategy comparisons.

    Mingxin YUAN, Sun-an WANG, Canyang WU, Kunpeng LI

    Inspired by the mechanism of Jerne’s idiotypic network hypothesis, a new adaptive immune network algorithm (AINA) is presented through the stimulation and suppression between the antigen and antibody by taking the environment and robot behavior as antigen and antibody respectively. A guiding weight is defined based on the artificial potential field (APF) method, and the guiding weight is combined with antibody vitality to construct a new antibody selection operator, which improves the searching efficiency. In addition, an updating operator of antibody vitality is provided based on the Baldwin effect, which results in a positive feedback mechanism of search and accelerates the convergence of the immune network. The simulation and experimental results show that the proposed algorithm is characterized by high searching speed, good convergence performance and strong planning ability, which solves the path planning well in complicated environments.

    Yongzhao LI, Leonard J, CIMINI JR, Nagee HIMAYAT

    The performance of Space-Time Block Coding (STBC) with co-channelMIMO interference is investigated. For an interference-limited environment, the closed-formexpressions for the probability density functions of the signalto-interference ratio are derived and applied to analyze the outage probability with three typical types of co-channel MIMO interferers: STBC, open-loop spatial multiplexing and closed-loop spatial multiplexing. Both theoretical analyses and simulation results show that the performance of STBC is independent of the MIMO modes used in the interfering links.

    Behrouz MAHAM, Mérouane DEBBAH, Are HJ?RUNGNES

    Due to the limited energy supplies of nodes, in many applications like wireless sensor networks energyefficiency is crucial for extending the lifetime of these networks. We study the routing problem for multihop wireless ad hoc networks based on cooperative transmission. The source node wants to transmit messages to a single destination. Other nodes in the network may operate as relay nodes. In this paper, we propose a cooperative multihop routing for the purpose of power savings, constrained on a required bit error rate (BER) at the destination. We derive analytical results for line and grid network topologies. It is shown that energy savings of 100% are achievable in line and grid networks with a large number of nodes for BER= 10-4 constraint at the destination.