Olufemi S. Ojo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Olufemi S. Ojo has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Mayowa Oyedepo Oyediran
Ajayi Crowther University
4 shared publications - Matthew Olusegun Adigun
University of Zululand
3 shared publications - Sunday Adeola Ajagbe
Department of Computer Engineering, Abiola Ajimobi Technical University, Ibadan 200255, Nigeria
3 shared publications - Olakekan Akinbosoye Okewale
Department of Computer Science, Ajayi Crowther University, Oyo, Nigeria
2 shared publications - Princewill Chima Obuzor
University of Salford
2 shared publications - Abidemi Emmanuel Adeniyi
Department of Information Systems, Faculty of Computer Science and Engineering, Obafemi Awolowo University, Ile-Ife 220101, Nigeria
1 shared publication
Latest Publications
- Open Access
- 10 November 2026
- Published by Bexford Publishing in British Journal of Contemporary Research
- Vol. 1 (6) , 070
Machine Learning for Pre-Test Malaria Diagnosis from Symptoms, Clinical Text and Speech: A PRISMA Systematic Review
In the absence of microscopy and rapid tests, malaria diagnosis is presumptive in many clinics in Africa, leading to overuse of drugs and masking other causes of fever. It has been proposed, based on pre-test data (e.g., symptoms, clinical notes, patient voice), to use standard ML models as triage tools. We searched PubMed and Scopus for studies from 2010 to 2026 that trained and evaluated standard ML models on pre-test clinical data to paint a high RQ1 for malaria diagnosis or on voice or clini...
- 10 November 2026
- Published by Bexford Publishing in British Journal of Contemporary Research
- Vol. 1 (6) , 070
Open AccessMachine Learning for Pre-Test Malaria Diagnosis from Symptoms, Clinical Text and Speech: A PRISMA Systematic Review
In the absence of microscopy and rapid tests, malaria diagnosis is presumptive in many clinics in Africa, leading to overuse of drugs and masking other causes of fever. It has been proposed, based on pre-test data (e.g., symptoms, clinical notes, patient voice), to use standard ML models as triage tools. We searched PubMed and Scopus for studies from 2010 to 2026 that trained and evaluated standard ML models on pre-test clinical data to paint a high RQ1 for malaria diagnosis or on voice or clini...
- Open Access
- 4 January 2026
- Published by International Journal of Management and Data Analytics in International Journal of Management and Data Analytics (IJMADA)
- Vol. 6 (1) , 283-296
Comparative Evaluation of Multilayer Perceptron, Random Forest, and Support Vector Machine for Sepsis Classification Using Longitudinal Structured Clinical Data
Sepsis classification from routinely collected clinical data is challenging because positive labels are uncommon, repeated observations from the same patient are correlated, and leakage-prone preprocessing can produce misleadingly high performance. This study re-evaluated multilayer perceptron (MLP), random forest (RF), and support vector machine (SVM) models using a leakage-controlled patient-level design. The source file contained 1,552,210 hourly observations from 40,336 patients; 2,932 patients (7.27%) were ever sepsis-positive. A reproducible stratified sample of 10,000 patients (382,387 hourly observations) was selected, preserving patient-level prevalence, and divided into 8,000 development and 2,000 independent hold-out patients with no patient overlap. Nine routinely available predictors were retained after a development-only missingness audit: heart rate, oxygen saturation, temperature, systolic blood pressure, mean arterial pressure, diastolic blood pressure, respiratory rate, age, and gender. Median imputation, missingness indicators, scaling, imbalance handling, hyperparameter tuning, and threshold selection were confined to development data and patient-grouped cross-validation. On the untouched hold-out cohort, RF achieved the highest ROC-AUC (0.671; 95% CI 0.635–0.706) and PR-AUC (0.041), with sensitivity 0.538 and specificity 0.716. RF significantly exceeded MLP in ROC-AUC (difference 0.0317, p=0.012), whereas RF and SVM did not differ significantly (p=0.054). Calibrated Brier scores were similar (0.0175–0.0176). Permutation importance and SHAP identified heart rate, respiratory rate, temperature, and blood-pressure measures as influential RF predictors. The models show moderate discrimination but low positive predictive value under severe class imbalance; therefore, they should be regarded as feasibility-level decision-support models requiring external, prospective, temporal, and fairness validation before clinical use.
- 4 January 2026
- Published by International Journal of Management and Data Analytics in International Journal of Management and Data Analytics (IJMADA)
- Vol. 6 (1) , 283-296
Open AccessComparative Evaluation of Multilayer Perceptron, Random Forest, and Support Vector Machine for Sepsis Classification Using Longitudinal Structured Clinical Data
Sepsis classification from routinely collected clinical data is challenging because positive labels are uncommon, repeated observations from the same patient are correlated, and leakage-prone preprocessing can produce misleadingly high performance. This study re-evaluated multilayer perceptron (MLP), random forest (RF), and support vector machine (SVM) models using a leakage-controlled patient-level design. The source file contained 1,552,210 hourly observations from 40,336 patients; 2,932 patients (7.27%) were ever sepsis-positive. A reproducible stratified sample of 10,000 patients (382,387 hourly observations) was selected, preserving patient-level prevalence, and divided into 8,000 development and 2,000 independent hold-out patients with no patient overlap. Nine routinely available predictors were retained after a development-only missingness audit: heart rate, oxygen saturation, temperature, systolic blood pressure, mean arterial pressure, diastolic blood pressure, respiratory rate, age, and gender. Median imputation, missingness indicators, scaling, imbalance handling, hyperparameter tuning, and threshold selection were confined to development data and patient-grouped cross-validation. On the untouched hold-out cohort, RF achieved the highest ROC-AUC (0.671; 95% CI 0.635–0.706) and PR-AUC (0.041), with sensitivity 0.538 and specificity 0.716. RF significantly exceeded MLP in ROC-AUC (difference 0.0317, p=0.012), whereas RF and SVM did not differ significantly (p=0.054). Calibrated Brier scores were similar (0.0175–0.0176). Permutation importance and SHAP identified heart rate, respiratory rate, temperature, and blood-pressure measures as influential RF predictors. The models show moderate discrimination but low positive predictive value under severe class imbalance; therefore, they should be regarded as feasibility-level decision-support models requiring external, prospective, temporal, and fairness validation before clinical use.
- 22 July 2025
- Published by Springer Nature
- p. 119-131
A Video-Based Gender Classification System Using White Shark Optimizer Based Support Vector Machine
- 22 July 2025
- Published by Springer Nature
- p. 119-131
A Video-Based Gender Classification System Using White Shark Optimizer Based Support Vector Machine
- Open Access
- 1 October 2024
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 36 (1) , 685-693
Comparative analysis of selected optimization algorithms for mobile agents’ migration pattern
Mobile agents are agents that can migrate from host-to-host to work in a heterogeneous network environment. A mobile agent can migrate from host-to-host in its plan with the statistics generated on each host through a route known as migration pattern. Migration pattern therefore is the route the agents use to travel within the plan from the first host to the last host. However, there is a need for a comparison between the commonly used optimization algorithms in developing migration patterns for mobile agents with respect to some evaluation metrics. In this paper, the three techniques firefly algorithm (FFA), honeybee optimization (HBO) and particle swarm optimization (PSO) were used for developing migration patterns for mobile agents and their comparison was done based on migration time, time complexity and network load as metrics. PSO is discovered to perform better in terms of network load with an average of 242.3905 bits per second (bps), time complexity with an average of 41.2688 number of nodes (n), and migration/transmission time with an average of 4.203462 seconds (s).
- 1 October 2024
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 36 (1) , 685-693
Open AccessComparative analysis of selected optimization algorithms for mobile agents’ migration pattern
Mobile agents are agents that can migrate from host-to-host to work in a heterogeneous network environment. A mobile agent can migrate from host-to-host in its plan with the statistics generated on each host through a route known as migration pattern. Migration pattern therefore is the route the agents use to travel within the plan from the first host to the last host. However, there is a need for a comparison between the commonly used optimization algorithms in developing migration patterns for mobile agents with respect to some evaluation metrics. In this paper, the three techniques firefly algorithm (FFA), honeybee optimization (HBO) and particle swarm optimization (PSO) were used for developing migration patterns for mobile agents and their comparison was done based on migration time, time complexity and network load as metrics. PSO is discovered to perform better in terms of network load with an average of 242.3905 bits per second (bps), time complexity with an average of 41.2688 number of nodes (n), and migration/transmission time with an average of 4.203462 seconds (s).
- Open Access
- 1 June 2024
- Published by Institute of Advanced Engineering and Science in International Journal of Electrical and Computer Engineering (IJECE)
- Vol. 14 (3) , 3244-3255
Comprehensive review of load balancing in cloud computing system
Load balancing plays a critical role in optimizing resource utilization and enhancing performance in cloud computing systems. As cloud environments grow in scale and complexity, efficient load balancing mechanisms become increasingly vital. This paper presents a comprehensive review of load balancing techniques in cloud computing systems, with a focus on their applicability, advantages, and limitations. The review encompasses both static and dynamic load balancing approaches, evaluating their effectiveness in addressing the challenges posed by cloud infrastructure, such as heterogeneity, scalability, and variability in workload demands. Furthermore, the review examines load balancing algorithms considering factors such as resource utilization, response time, fault tolerance, and energy efficiency. Additionally, the impact of load balancing on cloud performance metrics, including throughput, latency, and scalability, is analyzed. This review aims to provide insights into the state-of-the-art load balancing strategies and serve as a valuable resource for researchers, practitioners, and system designers involved in the development and optimization of cloud computing systems.
- 1 June 2024
- Published by Institute of Advanced Engineering and Science in International Journal of Electrical and Computer Engineering (IJECE)
- Vol. 14 (3) , 3244-3255
Open AccessComprehensive review of load balancing in cloud computing system
Load balancing plays a critical role in optimizing resource utilization and enhancing performance in cloud computing systems. As cloud environments grow in scale and complexity, efficient load balancing mechanisms become increasingly vital. This paper presents a comprehensive review of load balancing techniques in cloud computing systems, with a focus on their applicability, advantages, and limitations. The review encompasses both static and dynamic load balancing approaches, evaluating their effectiveness in addressing the challenges posed by cloud infrastructure, such as heterogeneity, scalability, and variability in workload demands. Furthermore, the review examines load balancing algorithms considering factors such as resource utilization, response time, fault tolerance, and energy efficiency. Additionally, the impact of load balancing on cloud performance metrics, including throughput, latency, and scalability, is analyzed. This review aims to provide insights into the state-of-the-art load balancing strategies and serve as a valuable resource for researchers, practitioners, and system designers involved in the development and optimization of cloud computing systems.