Sabah Mohammed

Real-Time Clinical Decision Support System with Data Stream Mining

This research aims to describe a new design of data stream mining system that can analyze medical data stream and make real-time prediction. The motivation of the research is due to a growing concern of combining software technology and medical functions for the development of software application that can be used in medical field of chronic disease prognosis and diagnosis, children healthcare, diabetes diagnosis, and so forth. Most of the existing software technologies are case-based data mining systems. They only can analyze finite and structured data set and can only work well in their early years and can hardly meet todays medical requirement. In this paper, we describe a clinical-support-system based data stream mining technology; the design has taken into account all the shortcomings of the existing clinical support systems.

Improving the classification performance of biological imbalanced datasets by swarm optimization algorithms

Classification which is a popular supervised machine learning method has many applications in computational biology, where data samples are automatically categorized into predefined labels with the aid of data mining. Often the training samples contain very few instances of interest (e.g., medical anomalies, rare disease in a population, and unusual syndromes, etc.), but many normal instances. Such imbalanced ratio of data distributions among the target labels hampers the efficacy of classification algorithms, because the induced model has not been trained with sufficient amount of instances of the interesting label(s), but overwhelmed with ordinary training records. Traditional remedies attempt to rebalance the data distributions of the target classes, by inflating the interesting instances artificially, reducing the majority of the common instances or a combination of both. Though the fundamental concept is effective, there is no clear guideline on how to strike a balance between fabricating the rare samples and reducing the norms, with the purpose of maximizing the classification accuracy. In this paper, an optimization model using different swarm strategies (Bat-inspired algorithm and PSO) is proposed for adaptively balancing the increase/decrease of the class distribution, depending on the properties of the biological datasets. The optimization is extended for achieving the highest possible accuracy and Kappa statistics at the same time as well. The optimization model is tested on five imbalanced medical datasets, which are sourced from lung surgery logs and virtual screening of bioassay data. Computer simulation results show that the proposed optimization model outperforms other class balancing methods in medical data classification.

HCX: A Distributed OSGi Based Web Interaction System for Sharing Health Records in the Cloud

WITH the maturity of Web Services and Enterprise Service Oriented Architectures (SOA), new delivery and Web interaction models are now demonstrating how services can be traded outside traditional ownership and provisioning boundaries. The value of SOA comes from having an architecture that readily accommodates change. The more your business changes, the more SOA pays for itself. However, the initial build-out of SOA, prior to business change or service sharing, is cost-ineffective. By incorporating cloud computing in SOA, the time to value is shortened because you leverage ‘other people’s work’ as well as saving on infrastructure cost by leveraging on demand cloud based infrastructure services. This article outlines a distributed Web interactive system for sharing health records on the cloud using distributed OSGi services and consumers, called HCX (Health Cloud eXchange). This system allows for different health record and related healthcare services to be dynamically discovered and interactively used by client programs running within a federated private cloud. A basic prototype is presented as proof of concept along with a description to the steps and processes involved in setting up the underlying infrastructure. Finally some future directions for using this infrastructure are illustrated.

Stream-based Biomedical Classification Algorithms for Analyzing Biosignals

Classification in biomedical applications is an important task that predicts or classifies an outcome based on a given set of input variables such as diagnostic tests or the symptoms of a patient. Traditionally the classification algorithms would have to digest a stationary set of historical data in order to train up a decision-tree model and the learned model could then be used for testing new samples. However, a new breed of classification called stream-based classification can handle continuous data streams, which are ever evolving, unbound, and unstructured, for instance--biosignal live feeds. These emerging algorithms can potentially be used for real-time classification over biosignal data streams like EEG and ECG, etc. This paper presents a pioneer effort that studies the feasibility of classification algorithms for analyzing biosignals in the forms of infinite data streams. First, a performance comparison is made between traditional and stream-based classification. The results show that accuracy declines intermittently for traditional classification due to the requirement of model re-learning as new data arrives. Second, we show by a simulation that biosignal data streams can be processed with a satisfactory level of performance in terms of accuracy, memory requirement, and speed, by using a collection of stream-mining algorithms called Optimized Very Fast Decision Trees. The algorithms can effectively serve as a corner-stone technology for real-time classification in future biomedical applications.

Developing a secure distributed OSGI cloud computing infrastructure for sharing health records

Cloud Computing has become an emerging computing paradigm which brings new opportunities and challenges to overcome. While the cloud provides seemingly limitless scalability and an alternative to expensive data center infrastructure, it raises new issues in regards to security and privacy as processing and storage tasks are handed over to third parties. This article outlines a Distributed OSGi (DOSGi) architecture for sharing electronic health records utilizing public and private clouds which overcomes some of the security issues inherent in cloud systems. This system, called HCX (Health Cloud eXchange), allows for health records and related healthcare services to be dynamically discovered and interactively used by client programs accessing services within a federated cloud. A basic prototype is presented as proof of concept along with a description of the steps and processes involved in setting up the underlying security services. Several improvements have been added to HCX including a Role-Based Single-Sign-On (RBSSO).

Evaluation of Stream Mining Classifiers for Real-Time Clinical Decision Support System: A Case Study of Blood Glucose Prediction in Diabetes Therapy

Earlier on, a conceptual design on the real-time clinical decision support system (rt-CDSS) with data stream mining was proposed and published. The new system is introduced that can analyze medical data streams and can make real-time prediction. This system is based on a stream mining algorithm called VFDT. The VFDT is extended with the capability of using pointers to allow the decision tree to remember the mapping relationship between leaf nodes and the history records. In this paper, which is a sequel to the rt-CDSS design, several popular machine learning algorithms are investigated for their suitability to be a candidate in the implementation of classifier at the rt-CDSS. A classifier essentially needs to accurately map the events inputted to the system into one of the several predefined classes of assessments, such that the rt-CDSS can follow up with the prescribed remedies being recommended to the clinicians. For a real-time system like rt-CDSS, the major technological challenges lie in the capability of the classifier to process, analyze and classify the dynamic input data, quickly and upmost reliably. An experimental comparison is conducted. This paper contributes to the insight of choosing and embedding a stream mining classifier into rt-CDSS with a case study of diabetes therapy.

Identifying Middlewares for Mashup Personal Learning Environments

The common understanding of e-learning has shifted over the last decade from the traditional learning objects portals to learning paradigms that enforces constructivism, discovery learning and social collaboration. Such type of learning takes place outside the formal academic settings (e.g., seminars or lectures) where a learning environment is created by using some kind of web application mashup tools. The use of these mashup tools moves the learning environment further away from being a monolithic platform towards providing an open set of learning tools, an unrestricted number of actors, and an open corpus of artifacts, either pre-existing or created by the learning process – freely combinable and utilizable by learners within their learning activities. However, collaboration, mashup and contextualization can only be supported through services, which can be created and modified dynamically based on middlewares to suit the current needs and situations of learners. This article identifies middlewares suitable for creating effective personal learning environment based on Web 2.0 mashup tools. This article also proposed a general framework for constructing such personal learning environments based on Ambient Learning realized by learning agents and the use of Enterprise Mashup servers.

Recent advances in metaheuristic algorithms: Does the Makara dragon exist?

Metaheuristic algorithms (MHs) have a long history that can be traced back to genetic algorithms and evolutionary computing in the 1950s. Since February 2008, with the birth of the Firefly algorithm, MHs started to receive attention from researchers around the globe. Variants and new species of MH algorithms have bloomed like sprouts after rain. However, the necessity for creating more new species of such algorithms is questionable. It can be observed that these algorithms are fundamentally made up of several widely used core components. By explaining these components, the underlying design for a collection of the so-called modern MH optimisation algorithms is revealed. In this paper, the core components in some of the more popular MH algorithms are reviewed, thereby debunking the myths of their novelty, and perhaps dampening claims that something really ‘new’ is invented simply by branding an MH search method with the name of another living creature. Counterintuitive experimentations have shown that by taking snapshots, anyone can show some improvements of an MH over another in some situation. Mixing certain components up indeed adds advantage over the original MH. The same goes to extending MH with slight functional modification. This work also serves as a general guideline and a reference for any algorithm architect who wants to create a new MH algorithm in the future.

Improvised methods for tackling big data stream mining challenges: case study of human activity recognition

Big data stream is a new hype but a practical computational challenge founded on data streams that are prevalent in applications nowadays. It is quite well known that data streams that are originated and collected from monitoring sensors accumulate continuously to a very huge amount making traditional batch-based model induction algorithms infeasible for real-time data mining or just-in-time data analytics. In this position paper, following a new data stream mining methodology, namely stream-based holistic analytics and reasoning in parallel (SHARP), a list of data analytic challenges as well as improvised methods are looked into. In particular, two types of decision tree algorithms, batch-mode and incremental-mode, are put under test at sensor data that represents a typical big data stream. We investigate whether and to what extent of two improvised methods—outlier removal and balancing imbalanced class distributions—affect the prediction performance in big data stream mining. SHARP is founded on incremental learning which does not require all the training to be loaded into the memory. This important fundamental concept needs to be supported not only by the decision tree algorithms, but by the other improvised methods usually at the preprocessing stage as well. This paper sheds some light into this area which is often overlooked by data analysts when it comes to big data stream mining.

Finding approximate solutions of NP-hard optimization and TSP problems using elephant search algorithm

A novel bio-inspired optimization algorithm called elephant search algorithm (ESA) has been applied to solve NP-hard problems including the classical traveling salesman problem (TS) in this paper. ESA emerges from the hybridization of evolutionary mechanism and dual balancing of exploitation and exploration. The design of ESA is inspired by the behavioral characteristics of elephant herds; hence, the name Elephant Search Algorithm which divides the search agents into two groups representing the dual search patterns. The male elephants are search agents that outreach to different dimensions of search space afar; the female elephants form groups of search agents doing local search at certain close proximities. By computer simulation, ESA is shown to outperform other metaheuristic algorithms over the popular benchmarking optimization functions which are NP-hard in nature. In terms of fitness values in optimization, ESA is ranked after Firefly algorithm showing superior performance over the other ones. The performance of ESA is most stable when compared to all other metaheuristic algorithms. When ESA is applied to the traveling salesman problem, different ratios of gender groups yield different results. Overall, ESA is shown to be capable of providing approximate solutions in TSP.