Incorporating feature selection method into support vector regression for stock index forecasting

Stock index forecasting is one of the most difficult tasks that financial organizations, firms and private investors have to face. Support vector regression (SVR) has become a popular alternative in stock index forecasting tasks due to its generalization capability in obtaining a unique solution. However, the major limitation of SVR is that it cannot capture the relative importance of independent variables to the dependent variable when many potential independent variables are considered. This study incorporates feature selection method and SVR for building stock index forecasting model. The proposed model uses multivariate adaptive regression splines (MARS), an effective nonlinear and nonparametric regression methodology, to identify important forecasting variables. The obtained significant predictor variables are then served as the inputs for the SVR model. Experimental results reveal that the obtained important variables from MARS can improve the forecasting performance of the SVR models. Moreover, the MARS results provide useful information about the relationship between the selected predictor variables and stock index through the obtained basis functions, important predictor variables and the MARS prediction function. Hence, the proposed stock index forecasting model can generate good forecasting performance and exhibits the capability of identifying significant predictor variables, which provide valuable information for further investment decisions/strategies.     

An Out-of-Sample Test for Nonlinearity in Financial Time Series: An Empirical Application

This paper employs a local information, nearest neighbour forecasting methodology to test for evidence of nonlinearity in financial time series. Evidence from well-known data generating process are provided and compared with returns from the Athens stock exchange given the in-sample evidence of nonlinear dynamics that has appeared in the literature. Nearest neighbour forecasts fail to produce more accurate forecasts from a simple AR model. This does not substantiate the presence of in-sample nonlinearity in the series.     

Automated generation of new knowledge to support managerial decision-making: case study in forecasting a stock market

Abstract: The deluge of data available to managers underscores the need to develop intelligent systems to generate new knowledge. Such tools are available in the form of learning systems from artificial intelligence. This paper explores how the novel tools can support decision‐making in the ubiquitous managerial task of forecasting. For concreteness, the methodology is examined in the context of predicting a financial index whose chaotic properties render the time series difficult to predict. The study investigates the circumstances under which enough new knowledge is extracted from temporal data to overturn the efficient markets hypothesis. The efficient markets hypothesis precludes the possibility of anticipating in financial markets. More precisely, the markets are deemed to be so efficient that the best forecast of a price level for the subsequent period is precisely the current price. Certain anomalies to the efficient market premise have been observed, such as calendar effects. Even so, forecasting techniques have been largely unable to outperform the random walk model which corresponds to the behavior of prices under the efficient markets hypothesis. This paper tests the validity of the efficient markets hypothesis by developing knowledge‐based tools to forecast a market index. The predictions are examined across several horizons: single‐period forecasts as well as multiple periods. For multiperiod forecasts, the predictive methodology takes two forms: a single jump from the current period to the end of the forecast horizon, and a multistage web of forecasts which progresses systematically from one period to the next. These models are first evaluated using neural networks and case‐based reasoning, and are then compared against a random walk model. The computational models are examined in the context of forecasting a composite for the Korean stock market.     

ProfARIMA: A profit-driven order identification algorithm for ARIMA models in sales forecasting

In forecasting, evolutionary algorithms are often linked to existing forecasting methods to optimize their input parameters. Traditionally, the fitness function of these search heuristics is based on an accuracy measure. In this paper, however, we combine forecasting accuracy with business expertise by defining a flexible and easily interpretable profit function for sales forecasting, which is based on the profit margin of a given product, the volume of its sales and the accuracy of the forecast. ProfARIMA is a new procedure that selects the lags of a Seasonal ARIMA model according to the profit of a model's forecasts by taking advantage of search heuristics. This procedure is tested on both publicly available datasets and a real-life application with datasets of The Coca-Cola Company in order to assess its performance, both in profit and accuracy. Three different evolutionary algorithms were implemented during this testing process, i.e. Genetic Algorithms, Particle Swarm Optimization and Simulated Annealing. The results indicate that ProfARIMA always performs at least equally to the Box–Jenkins methodology and often outperforms this traditional procedure. For The Coca-Cola Company, our new algorithm in combination with Genetic Algorithms even leads to a significantly larger profit for out-of-sample forecasts.     

The use of data mining and neural networks for forecasting stock market returns

It has been widely accepted by many studies that non-linearity exists in the financial markets and that neural networks can be effectively used to uncover this relationship. Unfortunately, many of these studies fail to consider alternative forecasting techniques, the relevance of input variables, or the performance of the models when using different trading strategies. This paper introduces an information gain technique used in machine learning for data mining to evaluate the predictive relationships of numerous financial and economic variables. Neural network models for level estimation and classification are then examined for their ability to provide an effective forecast of future values. A cross-validation technique is also employed to improve the generalization ability of several models. The results show that the trading strategies guided by the classification models generate higher risk-adjusted profits than the buy-and-hold strategy, as well as those guided by the level-estimation based forecasts of the neural network and linear regression models.     

An evolution-based approach with modularized evaluations to forecast financial distress

Due to the radical changing of the global economy, a more precise forecasting of corporate financial distress helps provide important judgment principles to decision-makers. Although financial statements reflect a firm's business activities, it is very challenging to discover critical information from these statements. Applying machine learning algorithms can be demonstrated to improve forecasting accuracy in predicting corporate bankruptcy. In this paper, we introduce an evolutionary approach with modularized evaluation functions to forecast financial distress, which allows using any evolutionary algorithm to extract the set of critical financial ratios and integrates more evaluation function modules to achieve a better forecasting accuracy by assigning distinct weights. To achieve a more precise predicting accuracy, the undesirable forecasting results from some modules are weeded out, if their predicting accuracies are out of the allowable tolerance range as learned from our mechanism.     

Financial ratings with scarce information: A neural network approach

On a wake of Basel II Accord in 2004, banks and financial institutions can build an internal rating system. This work focuses on Italian small firms that are more hard to judge because quite often financial data are not simply available. The aim of this paper is to propose a simulation model for assigning rating judgements to these firms, using poor financial information.The proposed model produces a simulated counterpart of Bureau van Dijk-K Finance (BvD) rating judgements. It is clear that there are problems when small firms must be judged because it is difficult to obtain financial data; indeed in Italy these enterprises must deposit the balance-sheet in reduced form. Suggested methodology is a three-layer process where each of them is formed by, respectively, one, two and four feed-forward artificial neural networks with back-propagation algorithm. The proposed model is a good solution for evaluating small firms with poor financial information but not only: the research underlines and supports the ability of artificial neural networks of learning and reproducing some aspects or some features or behaviours of reality.     

Purposive discovery of operators

This paper introduces theGenerate, Prune and Prove (GPP) methodology for discovering definitions of mathematical operators. GPP is a task within the IL exploration discovery system. We developed GPP for use in the discovery of mathematical operators with a wider class of representations than was possible with the previous methods by Lenat and by Shen. GPP utilizes thepurpose for which an operator is created to prune the possible definitions. The relevant search spaces are immense and there exists insufficient information for a complete evaluation of the purpose constraint, so it is necessary to perform a partial evaluation of the purpose (i.e., pruning) constraint. The constraint is first transformed so that it is operational with respect to the partial information, and then it is applied to examples in order to test the generated candidates for an operator's definition. In the GPP process, once a candidate definition survives this empirical prune, it is passed on to a theorem prover for formal verification. In this paper, we describe the application of this methodology to the (re)discovery of the definition of multiplication for Conway numbers, a discovery which is difficult for human mathematicians. We successfully model this discovery process utilizing information which was reasonably available at the time of Conway's original discovery. As part of this discovery process, we reduce the size of the search space from a computationally intractable size to 3468 elements.     

Bridging the divide in financial market forecasting: machine learners vs. financial economists

Financial time series forecasting is a popular application of machine learning methods. Previous studies report that advanced forecasting methods predict price changes in financial markets with high accuracy and that profit can be made trading on these predictions. However, financial economists point to the informational efficiency of financial markets, which questions price predictability and opportunities for profitable trading. The objective of the paper is to resolve this contradiction. To this end, we undertake an extensive forecasting simulation, based on data from thirty-four financial indices over six years. These simulations confirm that the best machine learning methods produce more accurate forecasts than the best econometric methods. We also examine the methodological factors that impact the predictive accuracy of machine learning forecasting experiments. The results suggest that the predictability of a financial market and the feasibility of profitable model-based trading are significantly influenced by the maturity of the market, the forecasting method employed, the horizon for which it generates predictions and the methodology used to assess the model and simulate model-based trading. We also find evidence against the informational value of indicators from the field of technical analysis. Overall, we confirm that advanced forecasting methods can be used to predict price changes in some financial markets and we discuss whether these results question the prevailing view in the financial economics literature that financial markets are efficient.     

Forecasting classification of operating performance of enterprises by ZSCORE combining ANFIS and genetic algorithm

Classification of operating performance of the enterprises is not only a hot issue emphasized by the management, but it is an important reference for investors too in their decision-making. Generally speaking, when predicting or analyzing business performance classification, most researchers adopt corporate financial early warning or credit-rating models, which pretty much use previous data and facts. Therefore, this paper brings about an alternative method to discriminate between excellent and poor business management, so as to take preventive measures prior to business crisis or bankruptcy. We collected the financial reports and financial ratios from the listed firms in mainland China and Taiwan as our samples to build up four kinds of forecasting models for business performance. The empirical results show that the hybrid model provides better classification forecasting capability than the other models, while the ANFIS model adjusted by genetic algorithm could effectively enhance the classification forecasting capability.     

改进的XGBoost模型在股票预测中的应用

随着时代的不断进步，人民生活水平日益提高。在解决温饱问题之余，有了可供投资的余财。越来越多的人将目光转向股市投资，为股市发展提供了资金条件。然而在纷繁复杂的股票市场，如何寻找最优股成为亟待解决的问题。这不仅是投资者单方面的困惑，也是股票预测领域中学者们所关心的重点。通过网格搜索算法对XGBoost模型进行参数优化构建GS-XGBoost的金融预测模型，并将该模型运用于股票短期预测中。分别以中国平安、中国建筑、中国中车、科大讯飞和三一重工2005年4月至2018年12月28日的每日收盘价作为实验数据。通过实验对比，相较于XGBoost原模型、GBDT模型以及SVM模型，GS-XGBoost模型在MSE、RMSE与MAE三个评价指标上都表现出较好的预测结果。从而验证，GS-XGBoost金融预测模型在股票短期预测中具有更好的拟合性能。     

A combination of artificial neural network and random walk models for financial time series forecasting

Properly comprehending and modeling the dynamics of financial data has indispensable practical importance. The prime goal of a financial time series model is to provide reliable future forecasts which are crucial for investment planning, fiscal risk hedging, governmental policy making, etc. These time series often exhibit notoriously haphazard movements which make the task of modeling and forecasting extremely difficult. As per the research evidence, the random walk (RW) is so far the best linear model for forecasting financial data. Artificial neural network (ANN) is another promising alternative with the unique capability of nonlinear self-adaptive modeling. Numerous comparisons of the performances of RW and ANN models have also been carried out in the literature with mixed conclusions. In this paper, we propose a combination methodology which attempts to benefit from the strengths of both RW and ANN models. In our proposed approach, the linear part of a financial dataset is processed through the RW model, and the remaining nonlinear residuals are processed using an ensemble of feedforward ANN (FANN) and Elman ANN (EANN) models. The forecasting ability of the proposed scheme is examined on four real-world financial time series in terms of three popular error statistics. The obtained results clearly demonstrate that our combination method achieves reasonably better forecasting accuracies than each of RW, FANN and EANN models in isolation for all four financial time series.     

A hybrid genetic algorithm and tabu search approach for post enrolment course timetabling

The post enrolment course timetabling problem (PECTP) is one type of university course timetabling problems, in which a set of events has to be scheduled in time slots and located in suitable rooms according to the student enrolment data. The PECTP is an NP-hard combinatorial optimisation problem and hence is very difficult to solve to optimality. This paper proposes a hybrid approach to solve the PECTP in two phases. In the first phase, a guided search genetic algorithm is applied to solve the PECTP. This guided search genetic algorithm, integrates a guided search strategy and some local search techniques, where the guided search strategy uses a data structure that stores useful information extracted from previous good individuals to guide the generation of offspring into the population and the local search techniques are used to improve the quality of individuals. In the second phase, a tabu search heuristic is further used on the best solution obtained by the first phase to improve the optimality of the solution if possible. The proposed hybrid approach is tested on a set of benchmark PECTPs taken from the international timetabling competition in comparison with a set of state-of-the-art methods from the literature. The experimental results show that the proposed hybrid approach is able to produce promising results for the test PECTPs.     

Incorporating complementary ratios in the analysis of financial statements

Ratios are routinely used for extracting information from accounting reports. However, ratios present only part of the information available and can be easily complemented provided that information-technology facilities are accessible. This study infers the functional form of the information discarded by ratios. Then it develops extensions of ratios incorporating the discarded information, showing examples of their use and discussing the benefits obtained. Extensions of ratios seem promising as a facility attached to computerized databases of accounting reports. The concepts developed here are a step toward a more technology-supported analysis of financial statements.     

An evolutionary algorithm with guided mutation for the maximum clique problem

Estimation of distribution algorithms sample new solutions (offspring) from a probability model which characterizes the distribution of promising solutions in the search space at each generation. The location information of solutions found so far (i.e., the actual positions of these solutions in the search space) is not directly used for generating offspring in most existing estimation of distribution algorithms. This paper introduces a new operator, called guided mutation. Guided mutation generates offspring through combination of global statistical information and the location information of solutions found so far. An evolutionary algorithm with guided mutation (EA/G) for the maximum clique problem is proposed in this paper. Besides guided mutation, EA/G adopts a strategy for searching different search areas in different search phases. Marchiori's heuristic is applied to each new solution to produce a maximal clique in EA/G. Experimental results show that EA/G outperforms the heuristic genetic algorithm of Marchiori (the best evolutionary algorithm reported so far) and a MIMIC algorithm on DIMACS benchmark graphs.     

Reconfiguring existing pollutant monitoring stations by increasing the value of the gathered information

The objective of this study is to provide a framework for relocating or reconfiguring existing pollution monitoring station networks by using feature selection and data mining techniques. This methodology enables a partial redesign based on the maximization of the available information that is gathered by the pollution networks by the optimal data mining technique. It also considers requirements of the decision makers, like potential target places, etc.Since this methodology is based on the quality of forecasting, it can also be useful for auditing and forecasting. A case study is included in this paper. In light of the prediction results, a new way to relocate the existing monitoring station is proposed.     

Combining high frequency data with non-linear models for forecasting energy market volatility

The popularity of realized measures and various linear models for volatility forecasting has been the focus of attention in the literature addressing energy markets’ price variability over the past decade. However, there are no studies to help practitioners achieve optimal forecasting accuracy by guiding them to a specific estimator and model. This paper contributes to this literature in two ways. First, to capture the complex patterns hidden in linear models commonly used to forecast realized volatility, we propose a novel framework that couples realized measures with generalized regression based on artificial neural networks. Our second contribution is to comprehensively evaluate multiple-step-ahead volatility forecasts of energy markets using several popular high frequency measures and forecasting models. We compare forecasting performance across models and across realized measures of crude oil, heating oil, and natural gas volatility during three qualitatively distinct periods: the pre-crisis period, the 2008 global financial crisis, and the post-crisis period. We conclude that the newly proposed approach yields both statistical and economic gains, while reducing the tendency to over-predict volatility uniformly during all the tested periods. In addition, the proposed methodology is robust to a substantial structural break induced by the recent financial crisis. Our analysis favors median realized volatility because it delivers the best performance and is a computationally simple alternative for practitioners.     

Ratio Selection for Classification Models

This paper is concerned with the selection of inputs for classification models based on ratios of measured quantities. For this purpose, all possible ratios are built from the quantities involved and variable selection techniques are used to choose a convenient subset of ratios. In this context, two selection techniques are proposed: one based on a pre-selection procedure and another based on a genetic algorithm. In an example involving the financial distress prediction of companies, the models obtained from ratios selected by the proposed techniques compare favorably to a model using ratios usually found in the financial distress literature.     

A Knowledge-Based Approach to Effective Document Retrieval

This paper presents a knowledge-based approach to effective document retrieval. This approach is based on a dual document model that consists of a document type hierarchy and a folder organization. A predicate-based document query language is proposed to enable users to precisely and accurately specify the search criteria and their knowledge about the documents to be retrieved. A guided search tool is developed as an intelligent natural language oriented user interface to assist users formulating queries. Supported by an intelligent question generator, an inference engine, a question base, and a predicate-based query composer, the guided search collects the most important information known to the user to retrieve the documents that satisfy users' particular interests. A knowledge-based query processing and search engine is devised as the core component in this approach. Algorithms are developed for the search engine to effectively and efficiently retrieve the documents that match the query.