Haya Al-Askar

Regularized dynamic self-organized neural network inspired by the immune algorithm for financial time series prediction

Abstract This paper presents a novel type of recurrent neural network, the regularized dynamic self-organized neural network inspired by the immune algorithm. The regularization technique is used with the dynamic self-organized multilayer perceptrons network that is inspired by the immune algorithm. The regularization has been addressed to improve the generalization and to solve the over-fitting problem. In this work, the average values of 30 simulations generated from 10 financial time series are examined. The results of the proposed network were compared with the standard dynamic self-organized multilayer perceptrons network inspired by the immune algorithm, the regularized multilayer neural networks and the regularized self-organized neural network inspired by the immune algorithm. The simulation results indicated that the proposed network showed average improvement using the annualized return for all signals of 0.491, 8.1899 and 1.0072 in comparison to the benchmarked networks, respectively.

Predicting financial time series data using artificial immune system-inspired neural networks

This paper investigates a set of approaches for the prediction of noisy time series data; specifically, the prediction of financial signals. A novel dynamic self-organised multilayer neural network based on the immune algorithm for financial time series prediction is presented, combining the properties of both recurrent and self-organised neural networks. In an attempt to overcome inherent stability and convergence problems, the network is derived to ensure that it reaches a unique equilibrium state. The accuracy of the comparative evaluation is enhanced in terms of profit earning; empirical testing used in this work includes normalised mean square error NMSE to evaluate forecast fitness and also evaluates predictions against financial metrics to assess profit generation. Extensive simulations for multi-step prediction in stationary and non-stationary time series were performed. The resulting forecast made by the proposed network shows substantial profits on financial historical signals when compared to various solely neural network approaches. These simulations suggest that dynamic immunology-based self-organised neural networks have a better ability to capture the chaotic movement in financial signals.

Training Neural Networks as Experimental Models: Classifying Biomedical Datasets for Sickle Cell Disease

This paper discusses the use of various type of neural network architectures for the classification of medical data. Extensive research has indicated that neural networks generate significant improvements when used for the pre-processing of medical time-series data signals and have assisted in obtaining high accuracy in the classification of medical data. Up to date, most of hospitals and healthcare sectors in the United Kingdom are using manual approach for analysing patient input for sickle cell disease, which depends on clinician’s experience that can lead to time consuming and stress to patients. The results obtained from a range of models during our experiments have shown that the proposed Back-propagation trained feed-forward neural network classifier generated significantly better outcomes over the other range of classifiers. Using the Receiver Operating Characteristic curve, experiments results showed the following outcomes for our models, in order of best to worst: Back-propagation trained feed-forward neural net classifier: 0.989, Functional Link Neural Network: 0.972, in comparison to the Radial basis neural Network Classifiers with areas of 0.875, and the Voted Perception classifier: 0.766. A Linear Neural Network was used as baseline classifier to illustrate the importance of the previous models, producing an area of 0.849, followed by a random guessing model with an area of 0.524.

A Genetic Analytics Approach for Risk Variant Identification to Support Intervention Strategies for People Susceptible to Polygenic Obesity and Overweight

Obesity is a growing epidemic that has increased steadily over the past several decades. It affects significant parts of the global population and this has resulted in obesity being high on the political agenda in many countries. It represents one of the most difficult clinical and public health challenges worldwide. While eating healthy and exercising regularly are obvious ways to combat obesity, there is a need to understand the underlying genetic constructs and pathways that lead to the manifestation of obesity and their susceptibility metrics in specific individuals. In particular, the interpretation of genetic profiles will allow for the identification of Deoxyribonucleic Acid variations, known as Single Nucleotide Polymorphism, associated with traits directly linked to obesity and validated with Genome-Wide Association Studies. Using a robust data science methodology, this paper uses a subset of the TwinsUK dataset that contains genetic data from extremely obese individuals with a BMI ≥ 40, to identify significant obesity traits for potential use in genetic screening for disease risk prediction. The paper posits an approach for methodical risk variant identification to support intervention strategies that will help mitigate long-term adverse health outcomes in people susceptible to obesity and overweight.

A Framework on a Computer Assisted and Systematic Methodology for Detection of Chronic Lower Back Pain Using Artificial Intelligence and Computer Graphics Technologies

Back pain is one of the major musculoskeletal pain problems that can affect many people and is considered as one of the main causes of disability all over the world. Lower back pain, which is the most common type of back pain, is estimated to affect at least 60 % to 80 % of the adult population in the United Kingdom at some time in their lives. Some of those patients develop a more serious condition namely Chronic Lower Back Pain in which physicians must carry out a more involved diagnostic procedure to determine its cause. In most cases, this procedure involves a long and laborious task by the physicians to visually identify abnormalities from the patient’s Magnetic Resonance Images. Limited technological advances have been made in the past decades to support this process. This paper presents a comprehensive literature review on these technological advances and presents a framework of a methodology for diagnosing and predicting Chronic Lower Back Pain. This framework will combine current state-of-the-art computing technologies including those in the area of artificial intelligence, physics modelling, and computer graphics, and is argued to be able to improve the diagnosis process.

Recurrent neural networks inspired by artificial Immune algorithm for time series prediction

This paper presents a novel Dynamic Self-Organised Multilayer Neural Network that can be used for prediction of noisy time series data. The proposed technique is based on the Immune Algorithm for financial time series prediction; combining the properties of both recurrent and self-organised neural networks. The network is derived to ensure that a unique equilibrium state can be achieved to overcome the known stability and convergence problems. Extensive simulations for multi-step prediction in stationary and non-stationary time series were performed. The resulting projection made by the proposed network shows substantial profits on financial historical signals when compared to other neural network approaches. These simulations have suggested that dynamic immunology-based self-organised neural networks have a better ability to capture the chaotic movement in financial signals.

The Utilisation of Dynamic Neural Networks for Medical Data Classifications- Survey with Case Study

Various recurrent neural networks have been utilised for medical data analysis and classifications. In this paper, the ability of using dynamic neural network to medicine related problems has been examined. Furthermore, a survey on the use of recurrent neural network architectures in medical applications will be discussed. A case study using the Elman, the Jordan and Layer recurrent networks for the classifications of Uterine Electrohysterography signals for the prediction of term and preterm delivery for pregnant women are presented.

Feature Analysis of Uterine Electrohystography Signal Using Dynamic Self-organised Multilayer Network Inspired by the Immune Algorithm

Premature birth is a significant worldwide problem. There is little understanding why premature births occur or the factors that contribute to its onset. However, it is generally agreed that early detection will help to mitigate the effects preterm birth has on the child and in some cases stop its onset. Research in mathematical modelling and information technology is beginning to produce some interesting results and is a line of enquiry that is likely to prove useful in the early prediction of premature births. This paper proposes a new approach which is based on a neural network architecture called Dynamic Self-organised Multilayer Network Inspired by the Immune Algorithm to analyse uterine electrohystography signals. The signals are pre-processed and features are extracted using the neural network and evaluated using the Mean Squared Error, Mean absolute error, and Normalized Mean Squared Error to rank their ability to discriminate between term and preterm records.

Evaluation of Advanced Artificial Neural Network Classification and Feature Extraction Techniques for Detecting Preterm Births Using EHG Records

Globally, the rate of preterm births is increasing and this is resulting in significant health, development and economic problems. Current methods for the early detection of such births are inadequate. However, there has been some evidence to suggest that the analysis of uterine electrical signals, collected from the abdominal surface, could provide an independent and easier way to diagnose true labour and detect when preterm delivery is about to occur. Using advanced machine learning algorithms, in conjunction with electrohysterography signal processing, numerous studies have focused on detecting true labour several days prior to the event. In this paper however, the electrohysterography signals have been used to detect preterm births. This has been achieved using an open dataset that contains 262 records for women who delivered at term and 38 who delivered prematurely. Several new features from Electromyography studies have been utilized, as well as feature-ranking techniques to determine their discriminative capabilities in detecting term and preterm records. Seven artificial neural network algorithms are considered with the results showing that the Radial Basis Function Neural Network classifier performs the best, with 85% sensitivity, 80% specificity, 90% area under the curve and a 17% mean error rate.

Regularized Dynamic Self Organized Neural Network Inspired by the Immune Algorithm for Financial Time Series Prediction

A novel type of recurrent neural network, the regularized Dynamic Self Organised Neural Network Inspired by the Immune Algorithm, is presented. The Regularization technique is used with the Dynamic self-organized multilayer perceptrons network that is inspired by the immune algorithm. The regularization has been addressed to improve the generalization and to solve the over-fitting problem. The results of an average 30 simulations generated from ten stationary signals are demonstrates. The results of the proposed network were compared with the regularized multilayer neural networks and the regularized self organized neural network inspired by the immune algorithm. The simulation results indicated that the proposed network showed better values in terms of the annualized return in comparison to the benchmarked networks.