Welcome to Research Commons

Research Commons is the University of Waikato's open access research repository, housing research publications and theses produced by the University's staff and students.

Communities in Research Commons

Select a community to browse its collections.

Now showing 1 - 2 of 2

Recent Submissions

Accuracy of machine learning models versus "hand crafted" expert systems – A credit scoring case study
(Elsevier, 2009) Ben-David, Arie; Frank, Eibe
Relatively few publications compare machine learning models with expert systems when applied to the same problem domain. Most publications emphasize those cases where the former beat the latter. Is it a realistic picture of the state of the art? Some other findings are presented here. The accuracy of a real world “mind crafted” credit scoring expert system is compared with dozens of machine learning models. The results show that while some machine learning models can surpass the expert system’s accuracy with statistical significance, most models do not. More interestingly, this happened only when the problem was treated as regression. In contrast, no machine learning model showed any statistically significant advantage over the expert system’s accuracy when the same problem was treated as classification. Since the true nature of the class data was ordinal, the latter is the more appropriate setting. It is also shown that the answer to the question is highly dependent on the meter that is being used to define accuracy.
Where confidence fails, training prevails: Effects of behavioural skills training on improving kaimahi confidence when using patient management systems for outreach immunisation services.
(The University of Waikato, 2026) Squire, Danielle; Blackmore, Tania; Carlson, Teah
In New Zealand, immunisation rates among tamariki Māori are consistently lower nationally at all recommended age milestones. Access to clear and timely information can facilitate engagement with whānau Māori. This study evaluated the effectiveness of Behavioural skills training (BST) to improve kaimahi use of a patient management system (PMS) – Indici. A Kaupapa Māori approach was utilised to engage with the organisation. Seven participants were recruited from a local health and social services provider. Consultation supported development of task analyses used for training. BST occurs in four phases: instruction, modelling, rehearsal, and feedback. Data were collected following a multiple baseline design with a single baseline session, BST intervention, and a follow-up session to assess for maintenance. All participants demonstrated improvement in task accuracy following the introduction of BST and maintained task accuracy in a post-training follow-up. Despite improvements across all participants, task adherence was impacted during training. This study highlights that BST is an effective, socially valid tool for training simple skills for accessing information in a PMS such as Indici. However, further research is needed to examine the use of BST for teaching complex computer systems in applied health settings.
An experimental evaluation on the effect of dynamic chairs on students' behaviour, movement, and academic achievement in mathematics
(The University of Waikato, 2026-05-04) Martin, Isaac; Anderson, Angelika; Starkey, Nicola J.
Traditional classroom chairs are often rigid and inflexible, resulting in uncomfortable static sitting. Conversely, dynamic chairs are designed to promote healthy movement and comfort, which is hypothesised to improve student engagement, reduce disruptive behaviour, and, in turn, improve academic achievement. Currently, there is a lack of research into the effects of dynamic chairs on students’ behaviour and outcomes compared to regular classroom chairs. This thesis comprises a series of studies that investigate the effect of BodyfurnFlex chairs, a new type of dynamic chair, on the behaviour, movement, achievement, and perceptions of students compared with traditional classroom chairs. In Study 1, I used a multiple baseline ABAB reversal design to investigate the effect that BodyfurnFlex chairs had on students’ behaviour and movement in comparison to regular classroom chairs. Additionally, a survey was used to assess students' perceptions of the chairs, while classroom environmental conditions were monitored as potential confounding variables. The results showed that BodyfurnFlex chairs significantly increased students’ on-task behaviour and reduced disruptive behaviours, including chair tipping and out-of-seat behaviour. Meanwhile, both in-chair and overall movement in the classroom increased when students were seated in BodyfurnFlex chairs. The majority of participants preferred BodyfurnFlex chairs, finding them more comfortable and believing they made it easier to complete their schoolwork. The environmental conditions in the classroom remained within the recommended levels throughout data collection, indicating they were not confounding variables in this study. In Study 2, I conducted a detailed analysis of the movement and environmental data collected in Study 1. This included analysing different aspects of the in-chair movement data, including displacement, acceleration, and rotation. Additionally, I examined whether BodyfurnFlex chairs could influence noise levels in classrooms and explored possible connections between movement, environmental, noise, and behavioural variables. The results showed that both in-chair displacement and acceleration significantly increased when students sat in BodyfurnFlex chairs. However, while displacement was not associated with students’ behaviour, acceleration was, suggesting that student behaviour is not related to how much students move but rather how consistently they move. Noise levels in the classroom had a significant negative correlation with on-task behaviour and decreased when students sat in BodyfurnFlex chairs. All environmental variables (CO2, temperature, and humidity) showed small, non-significant correlations with all other variables, indicating that they were not confounding factors. In Study 3, I investigated the effect of BodyfurnFlex chairs on students' academic achievement across multiple mathematics assessments using a pre-post intervention between-groups design. Additionally, focus group discussions were conducted to gain greater insight into teachers' and students’ perceptions of the chairs. The results showed no significant difference in test grades between students who used the BodyfurnFlex chairs and those who used regular chairs. However, several limitations around the results, including the testing/grading system and design, are discussed. Students reported far greater satisfaction with BodyfurnFlex chairs compared to the regular chairs, stating that they were more comfortable and functional, aided their ability to do their schoolwork, and felt they had a positive effect on their test results.
Gene selection from microarray data for cancer classification - A machine learning approach
(Elsevier, 2005) Wang, Yu; Tetko, Igor V.; Hall, Mark A.; Frank, Eibe; Facius, Axel; Mayer, Klaus F.X.; Mewes, Hans W.
A DNA microarray can track the expression levels of thousands of genes simultaneously. Previous research has demonstrated that this technology can be useful in the classification of cancers. Cancer microarray data normally contains a small number of samples which have a large number of gene expression levels as features. To select relevant genes involved in different types of cancer remains a challenge. In order to extract useful gene information from cancer microarray data and reduce dimensionality, feature selection algorithms were systematically investigated in this study. Using a correlation-based feature selector combined with machine learning algorithms such as decision trees, nave Bayes and support vector machines, we show that classification performance at least as good as published results can be obtained on acute leukemia and diffuse large B-cell lymphoma microarray data sets. We also demonstrate that a combined use of different classification and feature selection approaches makes it possible to select relevant genes with high confidence. This is also the first paper which discusses both computational and biological evidence for the involvement of zyxin in leukaemogenesis.
‘Returning to normal?’: Bourdieu, systems thinking, and online teaching post-COVID-19
(University of Newcastle, 2025) Richardson, Anthony; Herb, Annika; Lloyd, Clare; Richardson, Anthony; Corino, Maria-Teresa
The COVID-19 pandemic and resultant transition to online teaching was a profound shock for universities and their stakeholders (teachers, students, and administrators). Bourdieu’s concept of hysteresis (systemic rupture) describes how such disruptions provide an opportunity to question and potentially reshape elements of our collective social reality. However, his work also explores how our shared sense of what is ‘normal’ tends to be deeply ingrained and resistant to change. As the pandemic fades into the background of our collective memory, many tertiary teachers may be tempted to view tertiary teaching as a rubber ball which can simply bounce back to ‘the way things were’. But the Complex Adaptive Systems (CAS) which underpin our social reality do not behave like rubber balls. The Panarchy (Adaptive Cycle) heuristic offers a useful lens for understanding how the resilience of such systems lies in their ability to adapt when faced with challenges like the pandemic. Such disruptions can ripple through interconnected parts of the system, leading to changes which persist even after the initial crisis has passed. Such systems do not ‘bounce back’ unchanged after such shocks. The longing to return to pre-pandemic norms in university teaching is both personally understandable and sociologically explicable by reference to Bourdieu’s concepts of habitus and field. Yet the disruption caused by COVID may be precisely the kind of transformative event that the panarchy model captures well. The changes triggered by the pandemic may endure beyond its official end, reshaping systems in ways we did not anticipate. In this sense, our systems are not always resilient in the way we want them to be.