MUT Institutional Repository (IR)

MUT IR serves as the home for the intellectual output of MUT academic communities, and will ultimately include digital dissertations, faculty publications, digital special collections, open access publications, open educational resources and much more.

Browse Communities

Quick access to the main repository content types

9 communities

1. Journal Articles 1404 2. Books and Book Chapters 96 3. Theses & Dissertations 138 4. Conference/ Workshops /Seminar/ Proceedings 88 5. Institutional Memory 42 6. Speeches and Lectures 39 7. Policies / Reports / Statutes 3 8. Projects & Business plan 0 9. Government publications 0

Recent Submissions

Access status: Open Access ,
Revisiting Multilayer Perceptrons in the Era of Kolmogorov– Arnold Networks: An Empirical Comparison on Image Classification
(Journal of Future Artificial Intelligence and Technologies, 2026-11-01) Muia, Charles Munyao; Mutiso, Geoffry; Wanjau, Stephen Kahara
Multilayer perceptrons (MLPs) remain a widely used architecture in deep learning, whereas Kolmogorov–Arnold networks (KANs) have emerged as an alternative that represents nonlinear transformations through learnable univariate functions, commonly parameterized using splines. Despite growing interest in KANs, their performance relative to MLPs on high-dimensional image classification tasks remains insufficiently characterized, partly because differences in model capacity and training protocols complicate direct comparisons. This study presents a parameter-controlled empirical comparison of MLPs and KANs on CIFAR-10 across five capacity tiers ranging from 64K to 4M parameters. The model configurations are matched to within 2% in parameter count, while the training protocol is held consistent to facilitate an assessment of architectural differences across capacity levels. The results show that KANs achieve higher mean classification accuracy than MLPs at low-to-medium capacities (T₁–T₄), with gains of 0.3–1.1 percentage points that decrease as model capacity increases. At the highest capacity (T₅), MLPs marginally outperform KANs by 0.3 percentage points. However, MLPs consistently deliver 3.0–5.3 times higher accuracy per GPU-hour, supported by a 3.1–5.4-fold training-throughput advantage and 18–27 fewer convergence epochs. Per-class results further reveal differences in the categories where each architecture performs relatively well, with MLPs showing marginals advantages in classes with lower intra-class variability and KANs exhibiting gains in visually complex, higher-variability classes. Overall, the findings suggest that the relative advantages of MLPs and KANs depend on model capacity and computational constraints: KANs offer modest accuracy gains at lower capacities, whereas MLPs provide greater computational efficiency across all evaluated tiers. These results support a complementary assessment of the two architectures under the specific experimental conditions considered.
Access status: Open Access ,
Transformer-Based Hyperspectral Feature Learning Model for Early Detection of Wheat Yellow Rust
(CAAI Artificial Intelligence Research, 2026-06-30) Mutiso, Geoffry; Muia, Charles Munyao; Ndia, John
Wheat yellow rust is a major threat to global food security, causing yield losses of up to 70% if not detected early. Hyperspectral imaging enables pre-symptomatic detection, but extracting discriminative spectral–spatial features is challenging due to high dimensionality, redundancy, limited labelled data, and subtle disease signatures. This study proposes a unified transformer-based framework that integrates self-supervised masked autoencoder (MAE) pretraining with a SpectralFormer classifier. The MAE learns spectral–spatial representations by reconstructing masked spectral tokens, thereby providing effective initialisation for downstream supervised learning. The pretrained encoder is subsequently fine-tuned for disease classification, and the fully integrated model is evaluated on a real-world UAV-acquired hyperspectral dataset. Results show that the proposed MAE-SpectralFormer achieves 98.6% accuracy, 98.4% F1-score, and 0.957 Receiver Operating Characteristic - Area Under the Curve (ROC-AUC), outperforming the strongest Convolutional Neural Network (CNN) baseline (Inception-ResNet) by 5.4 percentage points in overall accuracy and 6.0 points in Rust-class F1-score, and exceeding the supervised SpectralFormer by 2.6 and 2.4 points, respectively. These findings demonstrate that self-supervised spectral–spatial learning improves early disease detection and offers a scalable, data-efficient solution for hyperspectral crop monitoring.
Access status: Open Access ,
On Multilingual Text Summarization: A Review of Model Architectures, Techniques, and Research Gaps
(International Journal of Data Informatics and Intelligent Computing (IJDIIC), 2026-10-15) Muia, Charles Munyao; Oirere, Aaron Mogeni; Wanjau, Stephen Kahara
The rapid expansion of multilingual digital content has intensified demand for automated text summarization systems that can operate across diverse languages. Yet the architectural and methodological foundations of this domain remain dispersed across individual studies. This paper presents a systematic literature review of multilingual text summarization research in Natural Language Processing, with particular attention to model architectures, multilingual enabling techniques, and existing research gaps. Following a PRISMA-guided review protocol, 47 studies published between 2017 and 2026 were identified, screened, and synthesized from major scholarly databases using predefined eligibility and quality assessment criteria. The analysis shows that the literature is dominated by four dominant architectural paradigms: multilingual sequence-to-sequence models, large language models, adapter-based models, and hybrid architectures, with a recent shift toward large language model-based approaches. Across these studies, multilingual capability is primarily achieved through pre-training and transfer learning, with support from parameter-efficient adaptation and data augmentation. The review further identified four persistent limitation categories: low-resource language underrepresentation; dataset and benchmark constraints; evaluation limitations dominated by ROUGE overreliance; and generalization and scalability challenges, manifested as the curse of multilinguality. These findings indicate that current progress in the field is largely driven by scaling existing models rather than resolving structural multilingual limitations, particularly those related to data availability and model architecture. This paper provides a consolidated view of architectural and methodological trends and offers a structured, informed basis for future research on equitable multilingual summarization, particularly for low-resource languages.
Access status: Open Access ,
Examining Kenya’s Tourist Destinations’ Appeal: the Perspectives of Domestic Tourism Market
(Journal of Tourism & Hospitality, 2012-09-30) Ndivo, Rayviscic Mutinda; Waudo, Judith N; Waswa, Fuchaka
Kenya’s tourism activity has always been centered on the coastal beaches and a few game parks despite of being endowed with a unique combination of tourist attractions spread throughout the country. Noting this skewed nature of tourism development, this study aimed at investigating the status of appeal of the different Kenya’s tourist regions from the perspective of the domestic market with the goal of examining the efficacy of the destination o s p i t t appeal enhancers. Data was collected using semi-structured questionnaires and analysed using mainly descriptive statistics. The study found out that while majority of Kenyans considered taking a holiday as being important to them, a significant majority of them were not be aware of most of the country’s touristic attractions and hence, frequented only a few of the country’s attractions, mainly the coastal beaches. Further, the study found out that the most important sources of travel information available to Kenyans were personal experience and information from significant others; travel marketers were considered as being the least significant source of information. The study thus, concluded that the limited scope of tourist activities available within most of Kenya’s attractions, coupled with the limited sources of travel information may actually be leading causes of the lack of awareness and popularity of most Kenya’s tourist attractions. This study thus, recommends a deliberate effort on the part of the tourism policy makers and marketers to enhance the appeal of all the country’s attractions and position them as destinations in their own right. This would call for the adoption of a tourism development strategy that not only seeks to promote the country as one destination with varied tourist attractions, but one that would adopt a bottom-up approach, primarily focusing on developing the individual tourist attractions as a way of ensuring a country-wide competitiveness.
Access status: Open Access ,
Application of destination choice model: Factors influencing domestic tourists destination choice among residents of Nairobi, Kenya
(Tourism Management, 2012-12) Ndivo, Rayviscic Mutinda; Mayaka, Melphon
The objective of this study was to assess the status of Kenya's tourist destinations and the factors that determine their choices among the residents of Nairobi. The study adopted a cross-sectional descriptive survey design that covered 118 respondents randomly selected from adult patrons at an up-market shopping mall in Nairobi. Data were collected using self-administered questionnaires. The Hypotheses were tested using chi-squares and Pearson Product–Moment Correlation at 95% confidence level. The results indicated that tourism in Kenya is almost exclusively centred on the south coast beaches and a handful of game reserves or national parks. The Kenya's domestic tourist market considered individual trait factors as being more significant in determining the choice of a holiday destination than the environmental factors. The findings further indicated that the factors that motivated Nairobi residents in the choice of domestic tourism destination in rank order include: knowledge and adventure; economic concerns; personal safety; destination information; travel arrangement; destination features; family and friends; leisure and relaxation; religious and cultural considerations and travel bragging. Statistical tests failed to reject the hypothesis that majority of Kenya's domestic tourist attraction areas do not fall within the evoked set of the destination choice model.