Philippe BORIBO KIKUNDA

Prof. Dr. Philippe BORIBO KIKUNDA

orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
30 September 2025 in Journal of Computing Theories and Applications
Verified
Journal Article
Open access
PKPhilippe BORIBO KIKUNDA
Philippe BORIBO KIKUNDA
IKIssa Tasho Kasongo
Issa Tasho Kasongo
TNThierry Nsabimana
Thierry Nsabimana
JNJérémie Ndikumagenge
Jérémie Ndikumagenge
LNLongin Ndayisaba
Longin Ndayisaba
JKJules Kala
Jules Kala
This study examines the application of Educational Data Mining (EDM) to predict the academic per-formance of first-year students at the Catholic University of Bukavu and the Higher Institute of Edu-cation (ISP) in the Democratic Republic of Congo. The primary objective is to develop a model that can identify at-risk students early, providing the university with a tool to enhance student support and academic guidance. To address the challenges posed by data imbalance (where successful cases outnumber failures), the study adopts a hybrid methodological approach. First, the SMOTE algorithm was applied to balance the dataset. Then, a stacking classification model was developed to combine the predictive power of multiple algorithms. The variables used for prediction include the National Exam score (PEx), the secondary school track (Humanities), and the type of prior institution (public, private, or religious-affiliated schools), as well as age and sex. The results demonstrate that this approach is highly effective. The model is not only capable of predicting success or failure but also of forecasting students' performance levels (e.g., honors or distinctions). Moreover, the use of the Apriori association rule mining algorithm allowed the identification of faculty-specific success profiles, transforming prediction into an interpretable decision-support tool. This research makes several significant contributions. Practically, it provides the University of Bukavu with a tool for student orientation and early risk detection. Methodologically, it illustrates the effectiveness of a combined approach to EDM in an African context. However, the study acknowledges certain limitations, including the non-public nature of the data and the geographical specificity of the sample. It therefore proposes avenues for future research, such as the integration of Explainable AI (XAI) techniques for more refined and transparent analysis of the results.

11 June 2025 in Journal of Computing Theories and Applications
Verified
Journal Article
Open access
PDF file
PKPhilippe BORIBO KIKUNDA
Philippe BORIBO KIKUNDA
JNJérémie Ndikumagenge
Jérémie Ndikumagenge
LNLongin Ndayisaba
Longin Ndayisaba
TNThierry Nsabimana
Thierry Nsabimana
In a context where students face increasingly complex academic choices, this work proposes a recommendation system based on Bayesian networks to guide new baccalaureate holders in their university choices. Using a dataset containing variables such as secondary school section, gender, type of school, percentage obtained, age, and first-year honors, we have constructed a probabilistic model capturing the dependencies between these characteristics and the option chosen. The data is collected at the Catholic University of Bukavu, the Official University of Bukavu, and the Higher Institute of Education of Bukavu, preprocessed and then used to learn the structure via the hill-climbing algorithm with the BIC score using R's bnlearn tool. The model enables us to estimate the probability that a candidate will choose a given stream, depending on their profile. The approach has been validated using metrics such as BIC, cross-validation, and bootstrap and offers a good compromise between interpretability and predictive performance. The results highlight the potential of Bayesian networks in constructing explainable recommendation systems in the field of academic guidance. The system produces orientation probability maps for each candidate, which can be used by enrollment service advisers, as well as an ordered list of options relevant to the candidate's profile. With a remarkable performance on a test sample of precision@k=0.85, recall@k=0.61, ndcg=0.8, and Map=0.88, it constitutes an effective lever for reducing the risk of being misdirected in universities in South-Kivu, in the Democratic Republic of Congo

18 August 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
David LUPANDE MwenebituDavid LUPANDE Mwenebitu
David LUPANDE Mwenebitu
DADemelash Areda
Demelash Areda
PAPascal Adroba
Pascal Adroba
FMFrancklin Mwamba
Francklin Mwamba
PKPhilippe BORIBO KIKUNDA
Philippe BORIBO KIKUNDA
GNGaylord AMANI NGABOYEKA
Gaylord AMANI NGABOYEKA
In May 2026, the Democratic Republic of the Congo declared an outbreak of Bundibugyo virus disease (BVD) caused by Bundibugyo ebolavirus (BDBV), for which no licensed vaccine or antiviral treatment was available. We analysed the epidemic trajectory through 14 July 2026 and developed conditional six-month projections. Aggregated data from official INSP/COUSP Situation Reports (14 May–14 July 2026), supplemented by WHO and UN reports, were used to describe cases, deaths, case-fatality ratios (CFRs), and geographic spread. Log-linear models estimated phase-specific growth rates and doubling times; logistic, Gompertz, and constant-incidence scenarios were used for projections, with model comparison, Bayesian sensitivity analysis, deterministic sensitivity bounds, CFR confidence intervals, and rolling-origin validation. Confirmed cases increased from 8 to 2,073, with 796 deaths (national CFR, 38.4%), while affected health zones increased from 3 to 45 across five provinces. Doubling time increased from 8.9 to 19.5 and 25.1 days, representing a 64.1% reduction in growth rate. Ituri accounted for 89.9% of cases, while North Kivu had a CFR of 58.7%. Six-month projections ranged from 2,624 cases under rapid containment to 3,942 under continued deceleration and 10,839 with persistent recent incidence. Although proportional growth slowed, transmission and geographic expansion remained active. The divergence between modelling approaches and systematic under-prediction in temporal validation indicate substantial forecast uncertainty; projections should therefore be interpreted as conditional planning scenarios rather than deterministic forecasts.

6 October 2026 in Viruses
Verified
Journal Article
Open access
PDF file
David LUPANDE MwenebituDavid LUPANDE Mwenebitu
David LUPANDE Mwenebitu
DADemelash Areda
Demelash Areda
PAPascal Adroba
Pascal Adroba
FKFranklin Mwamba Kande
Franklin Mwamba Kande
PKPhilippe BORIBO KIKUNDA
Philippe BORIBO KIKUNDA
GNGaylord AMANI NGABOYEKA
Gaylord AMANI NGABOYEKA
In May 2026, the Democratic Republic of the Congo declared an outbreak of Bundibugyo virus disease caused by Bundibugyo ebolavirus. We analysed aggregated official Situation Reports (SitReps) issued from 14 May through 14 July 2026 to describe the reported epidemic trajectory, geographic expansion, mortality, interval-specific growth, and conditional six-month projections. Confirmed cases increased from 8 to 2073 and confirmed deaths reached 796; 45 health zones in five provinces were affected. Average interval-specific growth rates generally declined over the observation period, although cumulative cases and the geographic footprint continued to increase. Gompertz and Richards curves described the observed cumulative trajectory better than logistic and exponential curves according to root mean square error (RMSE), the corrected Akaike information criterion (AICc), and the Bayesian information criterion (BIC); this comparison reflects in-sample fit only. Conditional six-month scenarios varied substantially across assumptions (approximately 2435 to 10,839 cumulative cases by 14 January 2027) and are contingent and highly uncertain; Gompertz fits showed underprediction of the subsequently reported total in four temporal checks. Reported case fatality ratios (CFRs) varied across provinces but are vulnerable to outcome delay, incomplete case ascertainment, small denominators, and retrospective reclassification. The results support continued surveillance and response planning but do not identify the causes of epidemic slowing or establish that transmission was controlled.