Anang Hudaya Muhamad Amin

One shot associative memory method for distorted pattern recognition

In this paper, we present a novel associative memory approach for pattern recognition termed as Distributed Hierarchical Graph Neuron (DHGN). DHGN is a scalable, distributed, and one-shot learning pattern recognition algorithm which uses graph representations for pattern matching without increasing the computation complexity of the algorithm. We have successfully tested this algorithm for character patterns with structural and random distortions. The pattern recognition process is completed in one-shot and within a fixed number of steps.

M-Learning Management Tool Development in Campus-Wide Environment

Mobile learning (M-Learning) integrates the current mobile and wireless computing technology with education primarily to enhance the effectiveness of the traditional learning process. One of the difficulties in implementing M-Learning is to deliver the content efficiently. This paper focuses on the development of M-Learning management tool in campus-wide environment using the Microsoft .NET infrastructure. User acceptance study was carried out to measure the feasibility of the M-learning management application development. The results of the study indicate a tendency towards the M-Learning acceptance in campus-wide environment. The main objective of the works that have been carried out is to develop a server-side M-Learning application using the Microsoft .NET infrastructure. The works that has been carried out act as preliminary work for future development of M-Learning application in campus-wide environment.

Single-Cycle Image Recognition Using an Adaptive Granularity Associative Memory Network

Pattern recognition involving large-scale associative memory applications, generally constitutes tightly coupled algorithms and requires substantial computational resources. Thus these schemes do not work well on large coarse grained systems such as computational grids and are invariably unsuited for fine grained environments such as wireless sensor networks (WSN). Distributed Hierarchical Graph Neuron (DHGN) is a single-cycle pattern recognising algorithm, which can be implemented from coarse to fine grained computational networks. In this paper we describe a two-level enhancement to DHGN, which enables it to act as a standard binary image recogniser. This paper demonstrates that our single-cycle learning approach can be successfully applied to denser patterns, such as black and white images. Additionally we are able to load-balance the pattern recognition processes, irrespective of the granularity of the underlying computational network.

Collaborative-Comparison Learning for Complex Event Detection Using Distributed Hierarchical Graph Neuron (DHGN) Approach in Wireless Sensor Network

Research trends in existing event detection schemes using Wireless Sensor Network (WSN) have mainly focused on routing and localisation of nodes for optimum coordination when retrieving sensory information. Efforts have also been put in place to create schemes that are able to provide learning mechanisms for event detection using classification or clustering approaches. These schemes entail substantial communication and computational overheads owing to the event-oblivious nature of data transmissions. In this paper, we present an event detection scheme that has the ability to distribute detection processes over the resource-constrained wireless sensor nodes and is suitable for events with spatio-temporal characteristics. We adopt a pattern recognition algorithm known as Distributed Hierarchical Graph Neuron (DHGN) with collaborative-comparison learning for detecting critical events in WSN. The scheme demonstrates good accuracy for binary classification and offers low-complexity and high-scalability in terms of its processing requirements.

Cloudlet-based cyber foraging framework for distributed video surveillance provisioning

Continuous monitoring activities in surveillance system generate massive amount of data to be transferred from Internet-of-Things (IoT) devices (i.e., cameras and sensors) to a centralized processing unit for analysis. However, several issues need to be addressed, including data migration over bandwidth limited and high latency communication networks, and the heterogeneous nature of data obtained from the surveillance system. The aim of this paper is to develop a scalable and lightweight intelligent distributed surveillance system that make use of an integrated framework of Internet-of-Things (IoT) and cloud computing. It leverages on the pervasiveness in IoT and ubiquitous property of cloud computing technology. This paper introduces the concept of bringing cloud closer to the IoT in order to address the resource poverty of IoT and the dependency for massive data transmission to distant cloud. For providing real-time on-site object detection, this paper instantiates a cloudlet on resource-rich nearby infrastructure which is connected to the IoT devices for distributed retrieval and processing of critical and sensitive data.

Parallel Pattern Recognition Using a Single-Cycle Learning Approach within Wireless Sensor Networks

Pattern recognition applications such as natural phenomena detection and structural health monitoring have been widely applied using wireless sensor networks. These applications involve large amount of data to be analysed, and thus incur high computational time and complexity. In this paper, we present a parallel associative memory-based pattern recognition algorithm known as distributed hierarchical graph neuron (DHGN). It is a single-cycle learning algorithm with in-network processing capability; able to reduce computational loads by efficiently disseminates recognition processes throughout the network. Hence, suitable to be deployed in wireless sensor networks. The results of the accuracy and scalability tests show that our system performs with high accuracy and remains scalable for increases in pattern size and the number of stored patterns. The response time for pattern recognition remains within milliseconds irrespective of the size of the network.

Hadoop in OpenStack: Data-location-aware cluster provisioning

Nowadays, cloud based analytics platforms are replacing traditional physical clusters due to the high efficiency it provides. Such cloud platforms runs Hadoop on virtual clusters with remotely attached storage. In cloud architecture with multiple geographically separated regions, virtual machines (VMs) belonging to a virtual cluster are placed randomly. In order to run MapReduce jobs, data have to be moved to the regions where the VMs reside to achieve data locality. In this paper, we propose a data-location aware virtual cluster provisioning strategy to identify the data location and provision the cluster near to the storage. The use of bio-inspired optimization algorithms are considered for optimizing the placements of VMs. Data location aware cluster provisioning reduces the network distance between storage and the virtual cluster, resulting in faster job completion times.

Holographic graph neuron

This article proposes the usage of Vector Symbolic Architectures for implementing Hierarchical Graph Neuron. The adoption of VSA representation maintains previously reported properties and performance characteristics of HGN and further makes it suitable for implementation in distributed wireless sensor networks of tiny devices.

Under the Cloud: A Novel Content Addressable Data Framework for Cloud Parallelization to Create and Virtualize New Breeds of Cloud Applications

Existing data management schemes in clouds are mainly based on Google File System (GFS) and MapReduce. Problems arise when data partitioning among numerous available nodes therein. This research paper explores new methods of partitioning and distributing data, that is, resource virtualization in cloud computing. Loosely-coupled associative computing techniques, which have so far not been considered for clouds, can provide the break through needed for their data management. Applications based on associative computing models can efficiently utilize the underlying hardware to scale up and down the system resources dynamically. In doing so, the main hurdle towards providing scalable partitioning and distribution of data in the clouds is removed, bringing forth a vastly superior solution for virtualizing data intensive applications and the system infrastructure to support pay on per-use basis.

A divide-and-distribute approach to single-cycle learning HGN network for pattern recognition

Distributed Hierarchical Graph Neuron (DHGN) is a single-cycle learning distributed pattern recognition algorithm, which reduces the computational complexity of existing pattern recognition algorithms by distributing the recognition process into smaller clusters. This paper investigates an effect of dividing and distributing simple pattern recognition processes within a computational network. Our approach extends the single-cycle pattern recognition capability of Hierarchical Graph Neuron (HGN) for wireless sensor networks into the more generic framework of computational grids. The computational complexity of the hierarchical pattern recognition scheme is significantly reduced and the accuracy is improved. The single-cycle learning capability, which develops within the HGN, shows better noisy pattern recognition accuracy when size of the clusters is adapted to pattern data. The scheme lowers storage capacity requirements per node and incurs lesser communication complexity while retaining HGNs response-time characteristics. Higher recall accuracy and scalability of the scheme is tested by storing large numbers of binary character patterns and heterogeneous binary images. The results show that the response-time remains insensitive to the number of stored pattern, the accuracy is improved, and the system resource requirements are significantly reduced.