Search

This paper investigates methods to evolve an automated agent that executes a niche trading stock strategy. Unlike trading strategies that seek to exploit broad market trends, we choose a very specific strategy on the assumption that it will be easier to learn, require less input data to do so, and more straightforward to evaluate the agents performance. In this case, we select a Low Price Recovery Strategy (LPRS), which involves picking stocks that have a high likelihood of quickly recovering after a steep, one day decline in share price. A series of intelligent agents are evolved through the use of a Genetic Programing approach. The inputs to our algorithms included traditional stock performance metrics, sentiment indicators available from online sources, and associated classification rules. The essential aspects of the research discussed include: identification of opportunities, feature selection and extraction, design of various genetic programs for evolving the agent, and testing approaches for the agents. We demonstrate that the evolved agent yields results outperform a randomized version of the LPRS and the benchmark Standard & Poor’s 500 (S&P500) stock market index.

A haplotype is a set of single nucleotide polymorphisms (SNPs) from a given chromosome, and provides valuable information about complex diseases. Current practices that the inferring of large scale SNP haplotypes from raw SNP data (genotypes) using computational approaches has gained a lot of attention, but it presents a grand challenges as it is inherently a NP-Hard problem. In this paper, we propose a heuristic approach, Genetic Algorithm (GA) model for the haplotypes inference method, based on the maximum-likelihood estimates of haplotype frequencies under the assumption of Hardy-Weinberg proportions. The goal of the genetic algorithm method is to obtain high prediction accuracy within a reasonable computing time. The performance of our model was evaluated on both simulated datasets and real datasets, and these results are promising, indicating that our model is a potential computational tool for haplotype inferences.

Background Genomic islands (GIs) are clusters of alien genes in some bacterial genomes, but not be seen in the genomes of other strains within the same genus. The detection of GIs is extremely important to the medical and environmental communities. Despite the discovery of the GI associated features, accurate detection of GIs is still far from satisfactory. Results In this paper, we combined multiple GI-associated features, and applied and compared various machine learning approaches to evaluate the classification accuracy of GIs datasets on three genera: Salmonella, Staphylococcus, Streptococcus, and their mixed dataset of all three genera. The experimental results have shown that, in general, the decision tree approach outperformed better than other machine learning methods according to five performance evaluation metrics. Using J48 decision trees as base classifiers, we further applied four ensemble algorithms, including adaBoost, bagging, multiboost and random forest, on the same datasets. We found that, overall, these ensemble classifiers could improve classification accuracy. Conclusions We conclude that decision trees based ensemble algorithms could accurately classify GIs and non-GIs, and recommend the use of these methods for the future GI data analysis. The software package for detecting GIs can be accessed at http://www.esu.edu/cpsc/che_lab/software/GIDetector/.