SU+ Digital Repository

SU+ is an online repository for the preservation and promotion of assorted digital content at Strathmore University

Off-Campus Access to restriced resources (including the ExamsBank) now requires registration using an @strathmore.edu email address

Authentication is NOT required for On-Campus Access to content

Photo by @inspiredimages

Communities in DSpace

Select a community to browse its collections.

Now showing 1 - 5 of 7

Conferences / Workshops / Seminars + 474
Documents and Proceedings of Conferences, Seminars, Workshops (and more) held at Strathmore University
Digital Archives 178
Assorted collections of resources covering various subject themes contributed by Faculty and Library Staff
Reports / Policies + 8
Public reports and policy documents
Research / Researchers / Publications 4367
Researcher Profiles / Conference presentations / Published research articles / Faculty and Corporate research outputs
Strathmore Heritage Collection 41
A digital chronicle of the History of the University presented through a mix of pictures, videos and digitized publications

Recent Submissions

Unlocking biomedical data for AI health research in Africa using GeneNetwork
(Strathmore University, 2025) Kilyungi, B. M.
Genetic data analysis is essential for understanding biological processes and diseases. GeneNetwork (GN), an open-source platform with over 20 years of genetic and phenotypic data, relies on a complex relational database. However, the data is currently difficult to access and manipulate due to its complex underlying structures, including around 80 cross-referenced Structured Query Language (SQL) tables and various file types. This dissertation aimed to address the limitations of the GeneNetwork2 SQL database in representing and querying graph-like biological data by transforming it into the Resource Description Framework (RDF). A self documenting Domain Specific Language (DSL) was developed using GNU Guile to automate the conversion of GN’s MariaDB SQL database into RDF triples. This involved defining ontologies, mapping SQL views to RDF, and storing the data in Virtuoso. The framework’s effectiveness was evaluated by comparing query performance and output quality between SQL and SPARQL. Results showed that RDF transformation significantly improved query efficiency and semantic richness. At a 99.9% confidence level, SPARQL queries exhibit statistically significant faster execution times than the equivalent SQL queries. Additionally, RDF’s structured representation enabled intuitive querying and better relationship discovery, as demonstrated in retrieving mouse species details and searching GeneRIF entries. In conclusion, transforming GN’s data into RDF made complex queries faster and enhanced its FAIR (Findable, Accessible, Interoperable, Reusable) properties, improving accessibility through semantic enrichment and interoperability with federated services for both human and machine agents. This transformation unlocks the full potential of the data, laying the groundwork for a more adaptable, AI-ready GN service and providing valuable insights for the broader application of RDF in biological and clinical data integration. KEYWORDS: Artificial Intelligence, Data Accessibility, Data Interpretation, GeneNetwork, Biological Data, Data Discovery, Resource Description Framework (RDF), Metadata
Auto insurance fraud detection using machine learning
(Strathmore University, 2025) Kimani, R. W.
Rising vehicle insurance fraud significantly undermines the profitability of insurers and unfairly increases premiums for honest policyholders. To combat this growing threat, advanced machine learning (ML) techniques offer a promising solution for detecting fraudulent claims with greater accuracy and efficiency. This study develops and evaluates an ML-based fraud detection system using rich claim datasets that capture policyholder details, vehicle specifications, and claim attributes such as accident history and claim values. Four ML algorithms—Logistic Regression, Random Forest, K-Nearest Neighbors (KNN), and XGBoost—were trained and assessed using key performance metrics, including accuracy, precision, recall, and F1-score. The results indicate that while Random Forest and XGBoost achieved high accuracy, they exhibited lower recall, making them less effective in identifying fraudulent claims. In contrast, KNN and Logistic Regression demonstrated superior recall, essential for minimizing undetected fraud. Further optimization through hyperparameter tuning and ADASYN resampling improved KNN’s recall to 0.52 and its AUC score to 0.71, while Logistic Regression maintained a recall of 0.60. Based on its balanced performance and interpretability, Logistic Regression was selected for deployment in a web-based fraud detection system. The study concludes that implementing ML-driven fraud detection can significantly reduce fraudulent payouts, streamline claims processing, and enhance customer satisfaction. Future research should explore the use of more recent datasets, deep learning techniques, and alternative resampling methods to further refine fraud detection accuracy. Expanding the model to include other types of insurance, such as life and health, could enhance its applicability across the industry.
Application of machine learning in establishing determinants of growth in the horticultural export sub-sector in Kenya
(Strathmore University, 2025) Odera, Y. J. D.
The export industry is key in any country’s economic growth (Furuoka, Harvey & Munir, 2019) of any country. The export industry plays a crucial role in a country’s economic growth, yet factors influencing its stability and contribution to economic development remain areas of concern. In Kenya, the horticultural sub-sector has experienced slow growth over the past decade, prompting questions about the key challenges affecting its performance. This study aimed to identify factors influencing the growth of horticultural exports and explore ways to enhance the industry’s contribution to employment, foreign exchange, and overall economic stability. Using an augmented gravity model, the study analyzed variables such as exchange rates, agricultural GDP, interest rates, climate, trade distance, and preferential trade policies to assess their impact on Kenya’s horticultural trade. The findings underscore the importance of monitoring climatic conditions, as they significantly affect production, labor force participation, and economic stability. The results highlight the predictive model’s economic significance in shaping GDP, employment, trade balance, and market growth. These insights can guide stakeholders including Policymakers and farmers in making strategic decisions regarding high-value horticultural products, market investments, and effective marketing strategies to drive revenue growth Key Terms: Gravity Model, Horticultural Product Code, predictive modelling
Analyzing and predicting urbanization patterns in Nairobi using geographic information systems and data mining techniques
(Strathmore University, 2025) Baariu, Y. M.
Urbanization, characterized by the expansion and development of towns due to population growth from both natural increase and migration, has profoundly reshaped cities worldwide. Over the past four decades, the proportion of people residing in urban areas more than doubled. According to the United Nations, projections indicated that by 2050, approximately 68% of the global population would be living in cities. This growth implied that nearly 2.5 billion additional people would reside in urban areas by 2050, with the majority of this expansion occurring in Asia and Africa. In Kenya, rapid urbanization resulted in various challenges, including the proliferation of informal settlements and inadequate infrastructure. Nairobi, as a major metropolis, experienced substantial urban growth, exacerbating these issues. This study sought to model urbanization patterns in Nairobi through the use of Geographic Information System (Geographic Information Systems (GIS)) and Remote Sensing, alongside the application of Data Mining techniques to forecast future urban expansion. The methodology involved analyzing Land Satellite (LANDSAT) 8 satellite imagery from 2014 and 2024, computing the Normalized Difference Built-up Index (Normalized Difference Built-up Index (NDBI)), and utilizing change detection techniques to assess urban development. The findings can provide crucial insights for policymakers and urban planners, facilitating sustainable urban development and addressing challenges associated with rapid urbanization. This research contributed to a broader understanding of urban dynamics in Kenya and presented a predictive framework for future urban planning efforts. Keywords: Urbanization, GIS, Predictive Modeling, Machine Learning
Effectiveness of anti – livestock rustling policies and strategies in Tiaty, Baringo County
(Strathmore University, 2025) Nasiuma, B. T.
Cattle rustling and banditry remain major threats to security in Kenya’s Arid and Semi-Arid Lands (ASALs), particularly in the North Rift region. Despite various government interventions, these problems persist. They continue to cause loss of lives, displacement of communities, destruction of property, and disruption of socio-economic activities, ultimately hindering development. This study evaluated the effectiveness of anti-rustling policies in Tiaty Constituency, Baringo County, and the factors influencing their implementation. It focused on four specific objectives: to assess the extent to which existing policies and strategies have reduced cattle rustling and improved security; to investigate the socio-economic factors driving cattle rustling and their impact on policy effectiveness; to examine the role of political governance in enforcing anti-rustling measures; and to explore how traditional practices and cultural beliefs affect the implementation and outcomes of these policies. The study was anchored on Social Conflict Theory and Social Disorganization Theory. A mixed-methods approach using a convergent parallel design was employed. Data were collected through questionnaires, interviews, and focus group discussions involving policymakers, law enforcement officers, community leaders, and residents. Quantitative data were analyzed using descriptive statistics and ANOVA, while qualitative data were subjected to thematic analysis. Findings showed that government-led actions such as security deployment, disarmament operations, curfews, and patrols had improved safety in the short term but were largely reactive and unsustainable. Arrests and prosecutions were rare, weakening their deterrent effect. On the other hand, community-led efforts like cattle branding and elder-led mediation were more impactful but faced obstacles, including political interference, corruption, and inadequate infrastructure. Political instability and weak governance structures further impeded policy effectiveness. Additionally, widespread poverty, illiteracy, and unemployment contributed to the persistence of rustling, while deeply rooted cultural beliefs continued to normalize the practice. The study concludes that a combined approach, merging government policies with grassroots initiatives is key to addressing cattle rustling effectively. Strengthening governance, enhancing law enforcement, and tackling the underlying socio-economic drivers are critical. Investments in infrastructure, promotion of alternative livelihoods, and meaningful community participation in policy-making processes are also vital. Future research should focus on developing integrated, sustainable strategies to combat cattle rustling across ASAL regions.