Big Data Analytics for Space Data Processing
Siti Sophiayati Yuhaniz | Siti Mariyam Shamsuddin
Space can generate lots of data that can be useful and meaningful to people. For example, remote sensing images from various satellites and aircraft e.g. Landsat, SPOT, communication data from various communication satellites e.g. Inmarsat and IntelSat, and, aircraft data collected by satellites. Most of these data are coming from different sources. How can these data be used together to solve our problems? Space data are in various formats, unstructured, and big in volumes…continue reading
Machine Learning Big Data Framework and Analytics for Big Data Problems
Shafaatunnur Hasan | Siti Mariyam Shamsuddin| Noel Lopes
Generally, big data computing deals with massive and high dimensional data such as DNA microrray data, financial data, medical imagery, satellite imagery and hyperspectral imagery. Therefore, big data computing needs advanced technologies or methods to solve the issues of computational time to extract valuable information without information loss…continue reading
The Role of Intelligent Web Pre-caching for Big Data Scale
Sarina Sulaiman | Siti Mariyam Shamsuddin | Ajith Abraham | Waleed Ali | Farhan Mohamed
Big data scale challenges are affecting of the intelligent Web service in link of workloads. When there is a huge of data, the latency problem occurred. Therefore, the overall progress will decrease in performance and interactivity be the impact. Normally, either Web caching, pre-fetching or pre-caching will be used in handling the latency issues. Web pre-caching is the temporary storage of Web objects for later retrieval. There are 3 significant advantages to Web pre-caching as shown in Figure 1…continue reading
E-learning Big Data Computing
Nor Bahiah Ahmad, Norazah Yusof, Siti Zaiton Mohd Hashim, Norsham Idris
Big Data is the term used for a collection of data-sets so large and complex that it becomes difficult to process using on-hand database management tools or traditional data processing applications. Big data has also reshaping e-learning, whereby with the advancement of the technology, e-learning has been implemented in various platform over the internet, such as mobile technologies, multimedia, social media and in cloud. The large data created by learners while learning online, the teaching materials deployed in e-learning, and the instructors teaching techniques have been collected and stored in the database or log files. The integration of data from multiple data sources can be analyzed to be used to predict student’s success, to predict student’s personality and to develop adaptive learning…continue reading
Classification of Class Imbalance Datasets of Big Data Scale
Aida Ali | Siti Mariyam Shamsuddin | Anca Ralescu
Big data is defined as a high volume of data with velocity and variety information which needs new means of data processing. Most data mining techniques have difficulties to accommodate such demand of space and time requirements for big data. Inevitably, real world data applications often possess imbalanced class distribution, where the observations from one class, which is often the most interesting/ important class, are largely outnumbered by the observations from the prevalence classes…continue reading
Wireless Monitoring for Big Data
Evizal Abdul Kadir | Siti Mariyam Shamsuddin
Development of Information and Communication Technology (ICT) management to apply in several of information systems, an infrastructure required high capacity of data for database, application and storage. Big Data is emerging from the real of science projects at web companies to help companies like telecommunication provider, banking, stock exchange, industries and so on, to understand exactly which customers are not satisfy with service and what processes caused the dissatisfaction and predict which customers are going to change…continue reading
Searching and Mining Large Databases
Naomie Salim
The global pharmaceuticals market is worth US$808 billion a year and expected to be worth more than US$1 trillion by the end of 2014. There are a huge amount of data involved from historical collections such as in-house databases, commercial databases such as ID Alert and MDDR, databases of agencies involved in the promotion and regulation of drug-related research such as Available Chemicals Directory, compound acquisition programmes where there is increasing number of companies synthesising compounds for purchase and also combinatorial chemistry where large numbers of compounds synthesised simultaneously…continue reading
Pattern Extraction in Spatio-Temporal High Dimensional Data
Siti Zaiton Mohd Hashim | Aina Musdholifah
The availability of huge amount of spatial-temporal data that has been collected over the years has greatly challenged the ability to digest the data and to gain useful knowledge that would otherwise lost. Spatial-temporal data has large size, because it contains large amounts of spatial information and temporal information further increases the data size. Thus, two crucial challenges in extracting pattern of the spatial-temporal data is first its high-dimensional characteristic and second is the unavailability of techniques to test and validate the mining and clustering results…continue reading
QR Code based Geometric Moment Invariant for Big Data Storage
Vazeerudeen Abdul Hameed | Siti Mariyam Shamsuddin
This research proposed a graph based algorithm to generate geometric moment invariant functions for Quick Response (QR) code in the domain of big data storage. The graph based algorithm has been found to suffer from high computational complexity. One major cause of this problem is that the algorithm generates too many graphs that produce zero moment invariant functions. Hence, we propose an algorithm to determine and eliminate the zero moment invariant generating graphs and thereby generate non-zero moment invariant functions with reduced computational complexity for QR Code system which is useful especially in Big Data issues due to its fast readability and greater data storage capacity…continue reading
Massive Training Artificial Immune Recognition System for Big Data Scale
Hang See Pheng | Siti Mariyam Shamsuddin
The emerging of medical images modalities has produced large scale of images to provide the detail visualisation of condition in internal organ. The processing of this big data scale become a challenging tasks due to the conventional computation capability. Generally, the machine learning approach is implemented in the computer aided system (CAD). Figure 1 shows the overview of lung nodules detection process using CAD system to assist the radiologists. The choice of machine learning algorithms is become crucial as substantial training always involved in the huge scale of medical images…continue reading
Human Action Invarianceness for Recognition in Big data Video Images
Nilam Nur Amir Sjarif | Siti Mariyam Shamsuddin| Siti Zaiton Mohd Hashim
When video generates on more than million of image frames with several of human action and generates large of feature vector representation, how we can standardize the representation for better accuracy? Therefore, our goal is to improve scheme of Human Action Recognition (HAR) with geometrical moment features for better feature representation through supervised invariant discretization…continue reading
Multimodal Biometric Identification:Determining True Identities through Big Data
Wong Yee Leng | Siti Mariyam Shamsuddin
Most biometric identification systems deployed in real-world applications are multimodal biometric identification, which refers to identify an individual based on his/her physiological and behavioral characteristics. These system has the capability to reliably distinguish between an authorised person and an imposter, since biometric characteristics are distinctive, cannot be forgotten or lost and the person to be authenticated is physically present at the point of identification. However, Multimodal biometric systems often encounter significant limitations due to the issues of Big Data (refer to Figure 1)…continue reading
Big Data for Social Media Analytics
Mohammed Zuhair | Siti MariyamShamsuddin| NorBahiah Ahmad
In 60 seconds: “168” million emails are sent, “694, 445” Google searches are done, “695,000” Facebook status are updated, “98,000” tweets are sent, “370,000” Skype calls are made, “6,600” Flicker images are uploaded, “1,500” new blog entries are posted, and more than “600” new YouTube videos are uploaded…continue reading
Ontology-based Cloud Service Discovery Using Intelligent Agents and its role in Big Data
Abdullah Marich Ali | Siti Mariyam Shamsuddin | Fathy Essa Albourae
Big data is a term used to describe any huge amount of structured, semi-structured and unstructured data that is becomes difficult to be processed using traditional data processing application. One of many challenges in big data is the searching (knowledge discovery) which is represent our topic research. Huge amount of structured and semi-structured information for cloud services with different attributes and cost are available. Finding the suitable service from the increasing numbers of cloud services that satisfy the user requirements such as performance, cost and security have become a big challenge. Therefore, an intelligent service discovery system is necessary for searching and retrieving appropriate services accurately and quickly…continue reading
Novel Hybrid ESNN Algorithms for Big Data Classification Problems
Abdulrazak Yahya | Siti Mariyam Shamsuddin | Haza Nuzly Abdull Hamid
There are many reasons for this choosing. Based on Schliebs, Defoin-Platel, and Kasabov (2009a),The most benefit reason of ESNN is its fast learning and ability of the trained network to incrementally learn new samples without retraining. Also, ESNN is widely applied in classification tasks, such as face recognition (Wysoski et al., 2006b), person authentication based on audiovisual information (Wysoski et al., 2006a), taste recognition (Soltic et al., 2008), ecologi- cal problems (Platel et al., 2009) and has achieved better results than traditional methods…continue reading
CAPSO: Centripetal Accelerated Particle Swarm Optimization
Zahra Behesti | Siti Mariyam Shamsuddin
Meta-heuristic search algorithms are developed to solve optimization problems. Such algorithms are appropriate for global searches because of their global exploration and local exploitation abilities. Swarm intelligence (SI) algorithms comprise a branch of meta-heuristic algorithms that imitate the behavior of insects, birds, fishes, and other natural phenomena to find solutions for complex optimization problems. In this study, an improved particle swarm optimization (PSO) scheme combined with Newton’s laws of motion, the centripetal accelerated particle swarm optimization (CAPSO) scheme is introduced…continue reading
Automatic White Matter Lesion Quantification and Visualization on MRI
Ong Kok Haur | Siti Mariyam Shamsuddin | Norzaini Rose
The increasing size of the ageing population worldwide naturally leads to greater occurrences of neurological disorders in the foreseeable future. With advancement in MRI technology, white matter lesion show prominently on Fluid attenuation inversion recovery (FLAIR) sequence. These quantitative lesion load finding is therefore essential for normal aging research…continue reading
A Linear Time Complexity of Breadth-First Search Using P System with Membrane Division
Einallah Salehi | Siti Mariyam Shamsuddin| Kourosh Nemati
One of the known methods for solving the problems with exponential time complexity such as NP-complete problems is using the brute force algorithms. Recently, a new parallel computational framework called Membrane Computing is introduced which can be applied in brute force algorithms. The usual way to find a solution for the problems with exponential time complexity with Membrane Computing techniques is by P System with active membrane using division rule…continue reading
Optimization Pigment Production from Agricultural Waste using Heuristic Optimization Algorithm
Siti Nurulasilah Suhaimi |Siti Mariyam Shamsuddin | Wan Azlina Ahmad
The world today, agricultural waste issues is a popular issues that frequently been discussed all around the world. Lots of efforts and actions have been done to increase the use of agricultural waste, towards making world Green. What means by “Green”? It can be described as environmental friendly and healthier ecological. The main reason for the need of using agricultural waste is to protect the environment…continue reading
Multi Modal Features for Content Based Image Retrieval
A.Rahman M.Alamin | Siti Mariyam Shamsuddin
Over the past decade, a huge set of information has been digitized , These huge collections of diverse media types are made accessible to computer users, the availability of information about a given topic, In current scenarios, it is difficult to locate accurate and relevant information.Similarly, Image retrieval also suffers from the huge size of information. A number of image retrieval techniques are developed to facilitate the task. Content based image retrieval is one most common technique that uses the content of an image to identify and retrieve the relevant images. In this study, we present an approach to extract color,texture and shape features of an image for Content-based image retrieval (CBIR)…continue reading
Palm Vein Recognition
Selma Elnasir | Siti Mariyam Shamsuddin
Palm vein recognition for large corporations like multi braches companies will have thousands of employees. When creating the biometrics system each user will have multiple to create the stored template in the database. This huge amount of data demands high storage capabilities…continue reading
Modified Massive Training Artificial Neural Network for Distinction between Nodule And Non-Nodule
Kourosh Nemati | Siti Mariyam Shamsuddin
The main concern is the process in which massive training artificial neural network are modified to distinction between nodule and non-nodule in conventional chest radiographs. Two hidden layers were applied in order to enhance the original massive training artificial neural network to reach a high level of accuracy. Using more than one hidden layer in the experiment causes the neural network convergence process to slow down, whereas applying two momentum terms make the network to converge quickly. Despite it takes longer for two hidden layers network to learn, the gradient values are substantially low, which makes the error surface smooth resulting in lower possibility of trapping in the local minima…continue reading
