Chien-Feng Huang

A hybrid stock selection model using genetic algorithms and support vector regression

In the areas of investment research and applications, feasible quantitative models include methodologies stemming from soft computing for prediction of financial time series, multi-objective optimization of investment return and risk reduction, as well as selection of investment instruments for portfolio management based on asset ranking using a variety of input variables and historical data, etc. Among all these, stock selection has long been identified as a challenging and important task. This line of research is highly contingent upon reliable stock ranking for successful portfolio construction. Recent advances in machine learning and data mining are leading to significant opportunities to solve these problems more effectively. In this study, we aim at developing a methodology for effective stock selection using support vector regression (SVR) as well as genetic algorithms (GAs). We first employ the SVR method to generate surrogates for actual stock returns that in turn serve to provide reliable rankings of stocks. Top-ranked stocks can thus be selected to form a portfolio. On top of this model, the GA is employed for the optimization of model parameters, and feature selection to acquire optimal subsets of input variables to the SVR model. We will show that the investment returns provided by our proposed methodology significantly outperform the benchmark. Based upon these promising results, we expect this hybrid GA-SVR methodology to advance the research in soft computing for finance and provide an effective solution to stock selection in practice.

A comparative study of probability collectives based multi-agent systems and genetic algorithms

We compare Genetic Algorithms (GAs) with Probability Collectives (PC), a new framework for distributed optimization and control. In contrast to GAs, PC-based methods do not update populations of solutions. Instead they update an explicitly parameterized probability distribution p over the space of solutions. That updating of p arises as the optimization of a functional of p. The functional is chosen so that any p that optimizes it should be p peaked about good solutions. The PC approach has deep connections with both game theory and statistical physics. We review the PC approach using its motivation as the information theoretic formulation of bounded rationality for multi-agent systems (MAS). It is then compared with GAs on a diverse set of problems. To handle high dimensional surfaces, in the PC method investigated here p is restricted to a product distribution. Each distribution in that product is controlled by a separate agent. The test functions were selected for their difficulty using either traditional gradient descent or genetic algorithms. On those functions the PC-based approach significantly outperforms traditional GAs in both rate of descent, trapping in false minima, and long term optimization.

Agent-Based Model of Genotype Editing

Evolutionary algorithms rarely deal with ontogenetic, non-inherited alteration of genetic information because they are based on a direct genotype-phenotype mapping. In contrast, several processes have been discovered in nature which alter genetic information encoded in DNA before it is translated into amino-acid chains. Ontogenetically altered genetic information is not inherited but extensively used in regulation and development of phenotypes, giving organisms the ability to, in a sense, re-program their genotypes according to environmental cues. An example of post-transcriptional alteration of gene-encoding sequences is the process of RNA Editing. Here we introduce a novel Agent-based model of genotype editing and a computational study of its evolutionary performance in static and dynamic environments. This model builds on our previous Genetic Algorithm with Editing, but presents a fundamentally novel architecture in which coding and non-coding genetic components are allowed to co-evolve. Our goals are: (1) to study the role of RNA Editing regulation in the evolutionary process, (2) to understand how genotype editing leads to a different, and novel evolutionary search algorithm, and (3) the conditions under which genotype editing improves the optimization performance of traditional evolutionary algorithms. We show that genotype editing allows evolving agents to perform better in several classes of fitness functions, both in static and dynamic environments. We also present evidence that the indirect genotype/phenotype mapping resulting from genotype editing leads to a better exploration/exploitation compromise of the search process. Therefore, we show that our biologically-inspired model of genotype editing can be used to both facilitate understanding of the evolutionary role of RNA regulation based on genotype editing in biology, and advance the current state of research in Evolutionary Computation.

An intelligent model for pairs trading using genetic algorithms

Pairs trading is an important and challenging research area in computational finance, in which pairs of stocks are bought and sold in pair combinations for arbitrage opportunities. Traditional methods that solve this set of problems mostly rely on statistical methods such as regression. In contrast to the statistical approaches, recent advances in computational intelligence (CI) are leading to promising opportunities for solving problems in the financial applications more effectively. In this paper, we present a novel methodology for pairs trading using genetic algorithms (GA). Our results showed that the GA-based models are able to significantly outperform the benchmark and our proposed method is capable of generating robust models to tackle the dynamic characteristics in the financial application studied. Based upon the promising results obtained, we expect this GA-based method to advance the research in computational intelligence for finance and provide an effective solution to pairs trading for investment in practice.

Tracking extrema in dynamic environments using a coevolutionary agent-based model of genotype edition

Typical applications of evolutionary optimization in static environments involve the approximation of the extrema of functions. For dynamic environments, the interest is not to locate the extrema but to follow it as closely as possible. This paper compares the extrema-tracking performance of a traditional Genetic Algorithm and a coevolutionary agent-based model of Genotype Editing (ABMGE). This model is constructed using several genetic editing characteristics that are gleaned from the RNA editing system as observed in several organisms. The incorporation of editing mechanisms provides a means for artificial agents with genetic descriptions to gain greater phenotypic plasticity. By allowing the family of editors and the genotypes of agents to co-evolve using the re-generation of editors as a control switch for environmental changes, the artificial agents in ABMGE can discover proper editors to facilitate the tracking of the extrema in dynamic environments. We will show that this agent-based model, together with a coevolutionary mechanism, is more adaptive and robust than the GA. We expect the framework proposed in this paper to advance the current state of research of Evolutionary Computation in dynamic environments.

Exploration of RNA editing and design of robust genetic algorithms

This paper presents our computational methodology using genetic algorithms (GA) for exploring the nature of RNA editing. These models are constructed using several genetic editing characteristics that are gleaned from the RNA editing system as observed in several organisms. We have expanded the traditional genetic algorithm with artificial editing mechanisms as proposed by (Rocha, 1997). The incorporation of editing mechanisms provides a means for artificial agents with genetic descriptions to gain greater phenotypic plasticity, which is environmentally regulated. Our first implementations of these ideas have shed some light into the evolutionary implications of RNA editing. Based on these understandings, we demonstrate how to select proper RNA editors for designing more robust GAs, and the results show promising applications to real-world problems. We expect that the framework proposed both facilitate determining the evolutionary role of RNA editing in biology, and advance the current state of research in genetic algorithms.

Assessment of Hypervisor and Shared Storage for Cloud Computing Server

This paper introduces the Open filer operation system to setup a network storage environment, and to build five different private small-cloud computing servers (PSCCS). Then by using the iSCSI storage technique through the TCP/IP protocol to explore the accessing assessments of the Open filer storage device for each feasible cloud computing system is undertaken. This includes comparing reading efficiency between storage device and cache. This storage technique can alleviate each physical hosts burden, and also becomes a storage method of shared resources. These assessments can provide optimized solutions to small and medium enterprises (SMEs), schools, and social groups in choosing a private small-cloud computing server. Experiment results of accessing assessment of the Open filer storage device shows that cloud server VMware vSphere V (ESX server) has the highest performance rate, but comes with high costs, alternatively free software Proxmox VE is the best choice among private small-cloud computing server.

A study of a hybrid evolutionary fuzzy model for stock selection

Stock selection has long been a challenging and important task in finance. Recent advances in machine learning and data mining are leading to significant opportunities to solve these problems more effectively. In this study, we aim at developing a methodology for effective stock selection using fuzzy models as well as genetic algorithms (GA). We first devise a stock scoring mechanism using fundamental variables and apply fuzzy membership functions to re-scale the scores properly. The scores are then used to obtain the relative rankings of stocks. Top-ranked stocks can thus be selected to form a portfolio. Furthermore, we employ GA for optimization of model parameters and feature selection for input variables to the stock scoring model. We will show that the investment returns provided by our proposed methodology significantly outperform the benchmark return. Based upon the promising results obtained, we expect this hybrid fuzzy-GA methodology to advance the research in soft computing for finance and provide an effective solution to stock selection in practice.

A comparative study of stock scoring using regression and genetic-based linear models

Stock selection has long been a challenging and important task in investment and finance. Researchers and practitioners in this area often use regression models to tackle this problem due to their simplicity and effectiveness. Recent advances in machine learning (ML) are leading to significant opportunities to solve these problems more effectively. In this paper, we present a comparative study between the traditional regression-based and ML-based linear models for stock scoring, which is crucial to the success of stock selection. In ML-based models, Genetic Algorithms (GA), a class of well-known search algorithms in the area of ML, is used for optimization of model parameters and selection of input variables to the stock scoring model. We will show that our proposed genetic-based method significantly outperforms the traditional regression-based method as well as the benchmark. As a result, we expect this genetic-based methodology to advance the research in machine learning for finance and provide an attractive alternative to stock selection over the regression-based approach.

Probability collectives multi-agent systems: a study of robustness in search

We present a robustness study of the search of Probability Collectives Multi-agent Systems (PCMAS) for optimization problems. This framework for distributed optimization is deeply connected with both game theory and statistical physics. In contrast to traditional biologically-inspired algorithms, Probability-Collectives (PC) based methods do not update populations of solutions; instead, they update an explicitly parameterized probability distribution p over the space of solutions by a collective of agents. That updating of p arises as the optimization of a functional of p. The functional is chosen so that any p that optimizes it should be p peaked about good solutions. By comparing with genetic algorithms, we show that the PCMAS method appeared superior to the GA method in initial rate of decent, long term performance as well as the robustness of the search on complex optimization problems.