Browsing by Author "Vella, A. D."

Now showing 1 - 2 of 2
  • No Thumbnail Available
    ItemOpen Access
    The application of metrics to the measurement of quality systems
    (Cranfield University, 1995-01) Williamson, A.; Vella, A. D.
    Quality system auditing has been a topic of much recent discussion but there has not been a commensurate amount of research on how the audit process can make more effective use of auditor time and improve the quality of their judgements. An approach certification bodies and their clients may adopt is to measure the quality system activities of the client, and use these measurements (or metrics) to improve their understand of the quality system. If these metrics were available to auditors, they could identify the strengths and weaknesses of the quality system and assist in deciding if the client complies with ISO 9000. This could make the audit process more cost-effective and focused. A detailed study of the current auditing process used by certification bodies, and a survey of auditors, identified the quality system activities that provide the most confidence that the company has a compliant quality system. The "quality loop" activities of internal audit/management review and corrective action were found to be both the most important activities in showing compliance and provide the basis of metrics that can inform the auditor about the state of health of these activities. Metrics capable of measuring the other quality system activities required by ISO 9000 are identified, and their effectiveness in monitoring the quality system is discussed. The research shows that metrics concerned with the quality loop can provide useful information to an auditor and can help them in reaching a judgement. It is also shown that the use of metrics is constrained by organisational and technical factors, such as the size of the organisation being measured and the correspondence between the activity being measured and the requirements of ISO 9000.
  • Thumbnail Image
    ItemOpen Access
    An investigation of genetic algorithms and genetic programming
    (1996-06) Kolcu, Ibrahim; Vella, A. D.
    There are many regression techniques that try and fit known models to sets of data. For them we assume the functional form of the model and use analytic or statistical techniques to find the values of any unknown model parameters. If the target model is thought to be of sufficiently complex a form, the above techniques (i.e. analytic and statistical techniques) may fail to provide the desired results and alternative methods have to be used. This is even more important if the underlying model is itself unknown. Genetic algorithms and genetic programming are two techniques that may help in the search for suitable models. Unfortunately, however, both of these techniques have themselves parameters that need to be specified and there are no clear guidelines to aid such choice. A number of other implementation issues are also open questions and in this thesis we look at a number of ways of implementing genetic algorithms and genetic programs to evaluate alternatives. Simple target models are used throughout most of this work so that the effects of changes to the method's parameters can be monitored. We look at how population size, crossover probability and mutation rate affect the speed of convergence of the genetic algorithm to an acceptable model. One of the most difficult aspects of genetic programming is the issue of the meaning of the offspring produced by crossover or mutation. Some systems arrange that any offspring that do not have meaning are removed from the population. Others ensure that no such offspring can arise. In this work we look at what might happen if we always impose a meaning on all possible offspring. In the genetic programming part of this work we look at two representations of our models. In the first we used a fixed length representation, whilst in the second we used a tree to represent each member of the population. We also look at a number of fitness functions. The commonest such functions are based upon errors between the model and the data. For our fitness functions we also use their correlation coefficient. We found that a strategy that starts by using correlation coefficient and then a fitness that combines both correlation coefficient and error worked better.