Chan Huah Yong

Time series prediction using adaptive association rules

Grid computing is formed by a large collection of interconnected heterogeneous and distributed system. One of the grid computing purposes is to share computational resources. The efficiency and effectiveness of resource utilization of a grid greatly depend on the scheduler algorithm. The scheduler is able to manage the grid resources more effectively if we able to predict and provide it with the future state of grid resources. Therefore, this paper proposes a model to perform time series prediction using adaptive association rules. This model uses the idea that if a segment of a repeatable time series pattern has occurred, it has the possibility that the following segments of the repeatable pattern appear. Data mining and pattern matching techniques are being applied to mine for repeatable time series patterns. This model has the ability to provide confident level for each prediction it made and perform continuous adaptation. A prototype of this model is being developed and tested with four test cases. These test cases are relatively simple because our work on this time series prediction using adaptive association rules is very much in its early stages. The result from the experiment shows that our model is able to capture repetitive time series patterns and perform prediction using those patterns. However, this model has some drawbacks such as it required high computational power and required large storage.

Architecture of agent-based healthcare intelligent assistant on grid environment

In this paper we present an architecture of agent-based Healthcare Intelligent Assistant on grid environment. The Healthcare Intelligent Assistant is used by the medical practitioners to retrieve and use various existing organizational knowledge to help solve medical cases. We will use case-based format that captures the experiential knowledge of healthcare practitioners [1]. Using grid technologies all resources-knowledge within hospitals can be use by medical practitioner’s across geographically distributed location.

A Market-Based Scheduler for JXTA-Based Peer-to-Peer Computing System

Peer-to-Peer (P2P) computing is said to be the next wave of computing after client-server and web-based computing. It provides an opportunity to harness a lot of idle peer-resources such as desktop computers across the Internet, for solving large-scale computing applications. Each peer is autonomous and it needs incentive for sustained contribution of its resources to P2P applications. In addition, a flexible and efficient job scheduling is needed to harvest the idle computing power as cheaply and economically as possible.

Harmony-based monarch butterfly optimization algorithm

Monarch butterfly optimization (MBO) is a new metaheuristic algorithm mimics the migration of butterflies from northern USA to Mexico. In MBO, there are mainly two processes. In the first process, the algorithm emulates how some of the butterflies move from the current position to the new position by the migration operator. In the latter process, the algorithm tunes the position of other butterflies by adjusting operator. In order to enhance the search ability of MBO, an innovation method called MBHS is introduced to tackle the optimization problem. In MBHS, the harmony search (HS) adds mutation operators to the process of adjusting operator to enhance the exploitation, exploration ability, and speed up the convergence rate of MBO. For the purpose to validate the performance of MBHS, 14 benchmark functions are used, and the performance is compared with well-known search algorithms. The experimental results demonstrate that MBHS performs better than the basic MBO and other algorithms.

A potential way for efficient information sharing based on mobile text messaging

These days most educated individuals own a mobile phone that is rapidly becoming one of their primary and main commutation devices. The most basic mobiles are able to send and receive text messages, which are normally considered to be the most energy efficient and cheapest commutation method. This paper was inspired by the simple communication method used in mobile devices, Short Message Service (SMS) or text messaging, creating an efficient geographical information sharing platform for ubiquitous and volunteer data gathering. With the help of two open SOA platforms 1) Skype cloud SMS-gateway, and 2) Sensor.Network a web of things platform, we were able to create a globally accessible information platform which is capable of receiving spatial information via SMS from anywhere anytime. This information platform is compatible with the Open Geographical Consortium (OGC) standards and all data can be accessed via the web services for further process and sharing purposes.

Cooperative Multi-Agent System for solving Packet World Problem in Grid

Multi-Agent System (MAS) is considered as a branch of distributed systems which has many applications such as E-commerce, transportation system, web application, and air traffic control. Packet World Problem is MAS application used in this research to study the cooperation interaction between agents. This research aims to prove that using cooperative MAS algorithm will solve packet world problem in minimum number of steps, less time, and decrease network consumption. This will minimize processor and memory utilization. Packet World Problem is a case study used in literature to study MAS interactions and characteristics. It is an environment consists of multiple agents and packets, whereas they were distributed on different Grid machines. The main goal of this problem is cleaning the environment from packets. The proposed work includes Cooperative MAS model which consists of two levels, where each level has a specific sub-goal to achieve. Agents at both levels cooperate to achieve the global goal. In the same time the proposed work includes Cooperative MAS algorithm which solve Packet World Problem efficiently. Different scenarios were tested and the proposed algorithm gave results as enhancement of 84% less in number of steps, 40% less in consumed time and 72% less in network consumption.

Resource Discovery Using PageRank Technique in Grid Environment

The grid deals with large scale and ever-expanding environment which contains million of users and resources. For this reason, resource selection has been a challenging task especially in meeting user’s demand for a quality of service (QoS). A quality of service is the ability to serve a job by providing quality and reliable resource in fulfilling the user’s need. Quality and reliable resource selections naturally yield excellent and quality results. The background of the users and where the resource belongs to are important in determining the quality of a resource. This paper concerns with efficient and quality-based resource discovery using Condor ClassAd and PageRank technique in order to achieve a quality resource matching. The paper discusses how quality of users and resources are determined and considered in the discovery process prior to allocating jobs to resources.

When Robots Meet Ubiquity and Grid Computing: Pervasive Grid Platform-Based Collaborative Work

Pervasive grid technology has emerged as future trend in research area, which induce new challenges and very promising merging of high performance and ubiquity. However, extending the group work in this new environment remains an important challenge to provide humans with comfortable and adequate pervasive features. Consequently, a given solution would lead to a developed grid support for collaborative work and, thereafter, would conduce to successful supply of enough facilities to reach the real environment. Currently, available pervasive grid solutions do not provide plenty of support to such required environment due to sensing restrictions, classical grid limitations for enabling ubiquity devices as part of heterogeneous resources and diverse constraints on collaborative developments especially related to the scalability of groupware. The purpose of this work is to promote a platform tying together high performance of group-work environment, transparent grid infrastructure dedicated to ubiquitous computing and commodious brace of collaborative groupware and finally diverse frameworks of interactions: human-machine interactions, human-human mediated interactions and human robot interactions. In this paper, we present eRoboRy platform, for: Collaborative Explorer Robots Platform for Remote Reality Discovery. It allows multiple robots, as intelligent devices, to operate collaboratively within a new structure of remote robot environment, which includes multiple operators as well. This research work undertakes to offer a new network of collaborative robots the ability to become thoroughly pervasive and to certify a large interactivity interfaces. This platform addresses the basic and new concept of PCRNet (Pervasive Collaborative Robot NETwork) based on PixRob (Pervasive mIXed reality environment of ROBot), as a remote environment of the deployed robots.

Multi-Agent Negotiation System Using Adaptive Fuzzy Logic in Resource Allocation

This paper addresses the multi-agent system (MAS) with negotiation capability using an artificial intelligent (AI) approach - adaptive fuzzy logic (AFL) - in learning the behavior of other agents when allocating resources. The learning continues during each reasoning process on the subsequent reaction of negotiants. We present the results of multi-agent negotiation with multi-issue (e.g. price, time, resource conditions) in three perspectives (accuracy, adaptability and reliability) with different AI and heuristic approaches. From our analysis, we found that the AFL gives an overall stable performance and outperform other methods in analyzing opponent tactics and strategies

Mapping search results into self-customized category hierarchy

With the rapid growth of online information, a simple search query may return thousands or even millions of results. There is a need to help user to access and identify relevant information in a flexible way. This paper describes a methodology that automatically map web search results into user defined categories. This allows the user to focus on categories of their interest, thus helping them to find for relevant information in less time. Text classification algorithm is used to map search results into categories. This paper focuses on feature selection method and term weighting measure in order to train an optimum and simple category model from a relatively small number of training texts. Experimental evaluations on real world data collected from the web shows that our classification algorithm gives promising results and can potentially be used to classify search results returned by search engines.