47 research outputs found
Fuzzy methodologies for automated University timetabling solution construction and evaluation
Get PDFThis thesis presents an investigation into the use of fuzzy methodologies for University timetabling problems. The first area of investigation is the use of fuzzy techniques to combine multiple heuristic orderings within the construction of timetables. Different combinations of multiple heuristic ordering were examined, considering five graph-based heuristic orderings - Largest Degree, Saturation Degree, Largest Enrolment, Largest Coloured Degree and Weighted Largest Degree. The initial development utilised only two heuristic orderings simultaneously and subsequent development went on to incorporate three heuristic orderings simultaneously. A central hypothesis of this thesis is that this approach provides a more realistic scheme for measuring the difficulty of assigning events to time slots than the use of a single heuristic alone. Experimental results demonstrated that the fuzzy multiple heuristic orderings (with parameter tuning) outperformed all of the single heuristic orderings and non-fuzzy linear weighting factors. Comprehensive analysis has provided some key insights regarding the implementation of multiple heuristic orderings.
Producing examination timetables automatically has been the subject of much research. It is generally the case that a number of alternative solutions that satisfy all the hard criteria are possible. Indeed, there are usually a very large number of such feasible solutions. Some method is required to permit the overall quality of different solutions to be quantified, in order to allow them to be compared, so that the best may be selected. In response to that demand, the second area of investigation of this thesis is concerned with a new evaluation function for examination timetabling problems. A novel approach, in which fuzzy methods are used to evaluate the end solution quality, separate from the objective functions used in solution generation, represents a significant addition to the literature.
The proposed fuzzy evaluation function provides a mechanism to allow an overall decision in evaluating the quality of a timetable solution to be made based on common sense rules that encapsulate the notion that the timetable solution quality increases as both the average penalty and the highest penalty decrease. New algorithms to calculate what is loosely termed the lower limits and upper limits of the proximity cost function for any problem instance are also presented. These limits may be used to provide a good indication of how good any timetable solution is. Furthermore, there may be an association between the proposed lower limit and the formal lower bound. This is the first time that lower limits (other than zero) have been established for proximity cost evaluation of timetable solutions
Fuzzy methodologies for automated University timetabling solution construction and evaluation
No full textThis thesis presents an investigation into the use of fuzzy methodologies for University timetabling problems. The first area of investigation is the use of fuzzy techniques to combine multiple heuristic orderings within the construction of timetables. Different combinations of multiple heuristic ordering were examined, considering five graph-based heuristic orderings - Largest Degree, Saturation Degree, Largest Enrolment, Largest Coloured Degree and Weighted Largest Degree. The initial development utilised only two heuristic orderings simultaneously and subsequent development went on to incorporate three heuristic orderings simultaneously. A central hypothesis of this thesis is that this approach provides a more realistic scheme for measuring the difficulty of assigning events to time slots than the use of a single heuristic alone. Experimental results demonstrated that the fuzzy multiple heuristic orderings (with parameter tuning) outperformed all of the single heuristic orderings and non-fuzzy linear weighting factors. Comprehensive analysis has provided some key insights regarding the implementation of multiple heuristic orderings.
Producing examination timetables automatically has been the subject of much research. It is generally the case that a number of alternative solutions that satisfy all the hard criteria are possible. Indeed, there are usually a very large number of such feasible solutions. Some method is required to permit the overall quality of different solutions to be quantified, in order to allow them to be compared, so that the best may be selected. In response to that demand, the second area of investigation of this thesis is concerned with a new evaluation function for examination timetabling problems. A novel approach, in which fuzzy methods are used to evaluate the end solution quality, separate from the objective functions used in solution generation, represents a significant addition to the literature.
The proposed fuzzy evaluation function provides a mechanism to allow an overall decision in evaluating the quality of a timetable solution to be made based on common sense rules that encapsulate the notion that the timetable solution quality increases as both the average penalty and the highest penalty decrease. New algorithms to calculate what is loosely termed the lower limits and upper limits of the proximity cost function for any problem instance are also presented. These limits may be used to provide a good indication of how good any timetable solution is. Furthermore, there may be an association between the proposed lower limit and the formal lower bound. This is the first time that lower limits (other than zero) have been established for proximity cost evaluation of timetable solutions
Automated examination timetabling solution construction using fuzzy approach
Get PDFExamination timetabling is of much interest and concern to academic institutions. The basic problem is to allocate a timeslot for all exams within a limited number of permitted timeslots in order to find a conflict free timetable. This assignment process is subject to ‘hard’ constraints which must be satisfied in order to get a feasible timetable, such as no student is required to sit two exams at the same time. In addition, it is also important to build a good quality examination timetable that considers not only the administration requirements, but also takes into account lecturers’ and students’ preferences. These requirements are generally considered as ‘soft’ constraints which are desirable (but not essential) to satisfy. As reported by Burke et al. (1996), these requirements vary from one academic institution to another
Development of a robust parser for extracting artifacts during model-based testing from uml diagrams
No full textDevelopment of a scalable parser for extracting artifacts during model-based test case generation process
No full textThe foundation of any software testing process
is test case generation. If test cases are poorly generated, model-based software testing results becomes unreliable or misleading. In model-based software testing, test cases are generated from any modeling tool that contains user’s requirements. To generate test cases from modeling tools, a parser is required to read, extract, interpret and display parsed requirements or artifacts which are executed to generate test cases. Existing model-based testing techniques lacks parsers that are capable of reading, extracting and interpreting all diagrams contained in a modeling tool at runtime. This has caused flexibility problems during test
case generation. Consequently, this research aims to
develop a scalable parser that can aid automatic test case generation from all UML diagrams at runtime
An approach for automatic generation of test cases from UML diagrams
No full textThe Unified Modeling Language (UML) is a standard notation used to model user's requirements for object oriented software systems. With the growing adoption of UML by software developers and researchers, studies have begun to investigate how it can be used to conduct testing. Several approaches of UML-based software testing have been proposed where test cases are derived from UML diagrams based on efficient algorithms. However, these approaches still suffer from three major limitations namely; inadequate test coverage criteria, insufficient diagrams coverages and incompatibility with various UML tools. Therefore, this paper proposes an improved approach for generating test cases from various UML diagrams based on full coverage criteria. To achieve this aim, a robust method for extracting artefacts from the underlying diagrams of the software under test (SUT) was developed, where the artefacts are represented in an intermediate form using a tree and test cases are generated by traversing the contents of the tree. The proposed approach is fully automated and the generated test cases satisfied the criteria defined for the generation process. The novelty of this approach is its systematic rather than ad hoc test case generation from UML diagrams to achieve high test coverag
Automatic generation of test cases from activity diagrams for UML based testing (UBT)
Get PDFActivity diagrams are one of UML behavioural models suitable for system testing because it has the capacity to effectively describe the behaviours of systems under development. In this paper, a technique is proposed that generates test cases from activity diagrams by constructing an activity flow tree (AFT) which stores all the information extracted from the model file of the diagram through the help of a parser. Then, we applied an algorithm to generate test cases from the constructed tree. Test cases were generated based on the elements of activity diagrams such as activity sequences, associated descriptions and conditions. The proposed technique generated accurate test cases that completely tallied with the modeled requirements in the diagram. We utilized all-paths, basic pair paths, conditions, branches and transition criteria for generating test cases using ATM withdrawal operation software as a case study
An automated approach based on bee swarm in tackling university examination timetabling problem
Get PDFA recently invented foraging behavior optimization algorithm which is the Artificial Bee Colony (ABC) algorithm has been widely implemented in addressing various types of optimization problems such as job shop scheduling, constraint optimization problems, complex numerical optimization problems, and mathematical function problems. However, the high exploration ability of conventional ABC has caused a slowdown in its convergence speed. Inspired from the Particle Swarm Optimization (PSO) method, an automated approach has been proposed in this study and is named as the Global Best Concept - Artificial Bee Colony (GBABC) algorithm. The algorithm is formulated using the global best concept, which is then implemented into the employed bee phase to incorporate the global best solution information into solutions. This is for the sake of leading the search process towards exploring other potential search regions to locate the best global solution. In addition, to improve its exploitation ability, a local search method has been incorporated into the onlooker bee phase. With the use of the global best concept and local search method, the convergence speed, exploration and exploitation abilities of the basic ABC have been significantly enhanced. Experiments are carried out on standard university examination benchmark problems (Carter’s un-capacitated dataset). Results obtained demonstrate that, generally, the GBABC had outperformed the basic ABC algorithm in almost all instances and its performance is also comparable to other published literature
