M. Nordin Zakaria

A Review on Particle Swarm Optimization Algorithm and Its Variants to Human Motion Tracking

Automatic human motion tracking in video sequences is one of the most frequently tackled tasks in computer vision community. The goal of human motion capture is to estimate the joints angles of human body at any time. However, this is one of the most challenging problem in computer vision and pattern recognition due to the high-dimensional search space, self-occlusion, and high variability in human appearance. Several approaches have been proposed in the literature using different techniques. However, conventional approaches such as stochastic particle filtering have shortcomings in computational cost, slowness of convergence, suffers from the curse of dimensionality and demand a high number of evaluations to achieve accurate results. Particle swarm optimization (PSO) is a population-based globalized search algorithm which has been successfully applied to address human motion tracking problem and produced better results in high-dimensional search space. This paper presents a systematic literature survey on the PSO algorithm and its variants to human motion tracking. An attempt is made to provide a guide for the researchers working in the field of PSO based human motion tracking from video sequences. Additionally, the paper also presents the performance of various model evaluation search strategies within PSO tracking framework for 3D pose tracking.

Agent Based Priority Heuristic for Job Scheduling on Computational Grids

Abstract A grid is an infrastructure for resource sharing. At present, many scientific applications require high computing power in processing, which can only be achieved by using the computational grid. For the selection and allocation of grid resources to current and future applications, grid job scheduling is playing a very vital role for computational grids. They constitute the building blocks for making grids available to the society. The efficient and effective scheduling policies, when assigning different jobs to specific resources, are very important for a grid to process high computing intensive applications. This paper presents an agent based job scheduling algorithm for efficient and effective execution of user jobs. This paper also includes the comparative performance analysis of our proposed job scheduling algorithm along with other well known job scheduling algorithms considering the quality of service parameters like waiting time, turnaround time, response time, total completion time, bounded slowdown time and stretch time. We also conducted the QoS based evaluation of the scheduling algorithms on an experimental computational grid using real workload traces. Experimental evaluation confirmed that the proposed grid scheduling algorithms posses a high degree of optimality in performance, efficiency and scalability. This paper also includes a statistical analysis of real workload traces to present the nature and behavior of jobs.

Comparative Performance of Modified Simulated Annealing with Simple Simulated Annealing for Graph Coloring Problem

The problems which are NP-complete in nature are always attracting the computer scientists to develop some heuristic algorithms, generating optimal solution in time-space efficient manner compared to the existing ones. Coloring of the vertices of a graph with minimum number of colours belongs to the same category, where the algorithm designers are trying to propose some new algorithms for better result. Here, we have designed modified Simulated Annealing (MSA) for optimal vertex coloring of a simple, symmetric and connected graph (GCP). The algorithm has been tested upon a series of benchmarks including large scale test case and has shown better output than the simple or non-modified version of the same algorithm. This paper describes the advancement of performance of simple SA applied upon the problem of graph coloring using a specially designed operator called random change operator instead of the general change operator. Our work is still going on for designing better algorithms generating optimal solutions.

Partially Optimized Cyclic Shift Crossover for Multi-Objective Genetic Algorithms for the multi-objective Vehicle Routing Problem with time-windows

The complexity of the Vehicle Routing Problems (VRPs) and their applications in our day to day life has garnered a lot of attentions in the area of optimization. Recently, attentions have turned to multi-objective VRPs with Multi-Objective Genetic Algorithms (MOGAs). MOGAs, thanks to its genetic operators such as selection, crossover, and/or mutation, constantly modify a population of solutions in order to find optimal solutions. However, given the complexity of VRPs, conventional crossover operators have major drawbacks. The Best Cost Route Crossover is lately gaining popularity in solving multi-objective VRPs. It employs a brute force approach to generate new children. Such approach may be unacceptable when presented with a relatively large problem instance. In this paper, we introduce a new crossover operator, called Partially Optimized Cyclic Shift Crossover (POCSX). A comparative study, between a MOGA based on POCSX, and a MOGA which is based on the Best Cost Route Crossover affirms the level of competitiveness of the former.

Stochastic partially optimized cyclic shift crossover for multi-objective genetic algorithms for the vehicle routing problem with time-windows

Graphical abstractDisplay Omitted HighlightsA stochastic partially optimized cyclic shift crossover (SPOCSX) is presented.BCRC and POCSX are used as reference crossovers to contrast the performances.Experiments show that SPOCSX produces higher quality solutions than POCSX.Experiments show that the execution time of SPOCSX is much lower than that of BCRC.Qualitative analysis shows the competitiveness of the solutions obtained by SPOCSX. This paper presents a stochastic partially optimized cyclic shift crossover operator for the optimization of the multi-objective vehicle routing problem with time windows using genetic algorithms. The aim of the paper is to show how the combination of simple stochastic rules and sequential appendage policies addresses a common limitation of the traditional genetic algorithm when optimizing complex combinatorial problems. The limitation, in question, is the inability of the traditional genetic algorithm to perform local optimization. A series of tests based on the Solomon benchmark instances show the level of competitiveness of the newly introduced crossover operator.

An Optimized Tuning of Genetic Algorithm Parameters in Compiler Flag Selection Based on Compilation and Execution Duration

Compiler flags exist to provide option for the software developer to dictate certain parameter to the compiler. Such parameters provide hints to the compiler on how to handle certain portion of the source code. In the realm of optimization, compiler flags provide the fastest way to speed up a program. The right combination of flags will provide significant enhancement in speed without compromising the integrity of the output. However, the main challenge is choosing that particular right set of flags. Many a times, developers work around this issue by dictating the optimization level. In that way, the compiler imposes a package of flags. This process may lead to degradation of performance in terms of execution speed and also significant increase in program size. In this work, we are studying the usage of Genetic Algorithm as a way to select the optimization flags that could produce codes which compile and execute fast.

Association among independent and dependent factors of host in volunteer computing

These days, many scientific projects such as: SETI@home and Distributed.net are using computers which are connected to internet from the entire world in order to run their jobs. This large distributed system is a Volunteer Computing (VC) that includes volunteer hosts. An important attribute of volunteer hosts is the volatile behavior that means often they are not available because of the autonomy nature of owners. The volatile behavior of hosts is significant and effective issue on fault tolerance of the system. Therefore, characterization factors of volatile hosts have influence on reliability of jobs schedulers and improve fault tolerance of the VC system. This paper considered on characterization hosts factors in concept of association between dependent and independent factors of hosts. We assumed that CPU interval availability and frequency number of CPU availability are dependent factors, also time zone, location, RAM size and numbers of processors are independent factors. In other words, the main hypothesis is that there is an association between dependent and independent factors. To proof this hypothesis, we used real trace data of volunteers which are taken from SETI@home project. We found that time zone and location have dependency, also processors number have correlation with CPU interval availability while frequency number of CPU availability has correlation with RAM size.

An efficient energy two mode error correction technique in underwater wireless sensor networks

Research in underwater acoustic networks has developed quickly to support large variety of applications such as mining equipment and environmental monitoring. As in terrestrial sensor networks; reliable data transport is demanded in underwater sensor networks. The energy efficiency of error correction techniques should be considered because of the severe energy constraints of underwater wireless sensor networks. FEC and ARQ are the two main error correction techniques in underwater networks. In this paper, a mathematical energy efficiency analysis for FEC and ARQ techniques in underwater environment is done based on communication distance, packet size and wind speed. And a comparison between FEC and ARQ in terms of energy efficiency is done, it was found that energy efficiency of both techniques increases with increasing packet size in short distances, but decreases in longer distances. There is also a cutoff distance below which ARQ is more energy efficient than FEC, and after which FEC is more energy efficient than ARQ. Wind speed has great effect on energy efficiency and on cutoff distance for both techniques. Based on this analysis an efficient energy two mode error correction technique is proposed, this technique depends on RTT, distance, packet size, and the energy level in the nodes to determine where and how error can be corrected.

Optimizing BOINC scheduling using genetic algorithm based on thermal profile

Berkeley Open Infrastructure for Network Computing (BOINC) is an open source middleware for volunteer and grid computing. Main function of BOINC is to use the idle time of computer to run some computation at background. Universiti Teknologi Petronas (UTP) campus grid used BOINC as middleware in computer labs. However, computer can only process jobs during weekday and office hour because they want to reduce energy used for cooling power. In order to fully utilize the computer in labs, we proposed new jobs scheduling algorithm can run based on thermal constraint. The proposed algorithm is combination of thermal profile and heuristic approach. We use genetic algorithm to find the best combination of clients and jobs based on clients order and least execution time. Then we compare our algorithm with brute force method. Result from simulation it shows that proposed algorithm successfully distribute and execute job based on thermal constraints in an effective and efficient way compare to brute force method.

Self-Organizing Map approach to determining compromised solutions for multi-objective UAV path planning

The demand for Unmanned Aerial Vehicle (UAV) extends to various civil and military missions. While the use of remotely controlled UAV reduces the rate of human casualties in hazardous environments, it is reported that most of UAV accidents are caused by human factor errors. Automated path planning is required and because of the multi-objective nature of UAVs missions, several heuristic approaches to path planning have been proposed in order to automate UAVs navigation. While solving multi-objective problems requires the search for a set of pareto-optimal points, it requires the involvement of the user to select the desired result from the solution space. In this paper, we propose a variant of Self-Organizing Map approach to finding a compromised solution for a multi-objective path planning problem that does not require user involvement. Preliminary tests conducted in virtual environments have shown the immunity of our algorithm to local minima, and its efficiency to respond to multiple objectives.