Mahamat Ali Issaka
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Mahamat Ali Issaka has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Ali Souleymane Dabye
Université Gaston Berger, UFR SAT, LERSTAD, Saint–Louis, Sénégal
2 shared publications - Lebede Ngartera
Department of Applied Mathematics, AI & Data Science, TeraSystemsAI LLC, Philadelphia, PA 19104, USA
2 shared publications - Saralees Nadarajah
Department of Mathematics, University of Manchester, Manchester M13 9PL, UK
2 shared publications - Ali Souleyman Dabye
UFR SAT, LERSTAD, Université Gaston Berger, Saint-Louis, Senegal, Senegal
1 shared publication - Hassan Goudja Izaddine
Université Adam Barka, Abéché, Chad, Chad
1 shared publication - Hissein Abba Annama
Laboratoire de Modélisation Mathématique, Informatique, Applications et Simulations (L2MIAS), Université de N’Djamena, BP 1027, N’Djamena, Chad, Chad
1 shared publication
Latest Publications
- Open Access
- 18 November 2026
- Published by MDPI AG in Preprints.org
Assessing Markov Persistence Beyond Seasonality in Monthly Rainfall-State Prediction in Chad
Does the previous month’s rainfall state improve forecasts once the calendar month is known? We used monthly rainfall records from Chad’s national meteorological agency (ANAM) for 2010–2020. The records cover 60 stations and 7,107 of 7,920 possible station-months. We analyzed four historical stations and extended the comparison to 30 stations selected for completeness, including the original four. An event was a monthly accumulation ≥ 20 mm. We compared calendar-month frequencies, homogeneous and two-season Markov chains, and seasonal logistic regressions with and without the previous state. Across 1,735 eligible rolling-origin forecasts in 2016–2020, pooled Brier scores were 0.05755 for calendar-month climatology, 0.07918 for the two-season Markov chain and 0.05789 for harmonic regression with a lag. The two-season chain outperformed climatology at one station. Adding the lag could improve a given seasonal regression, but the comparison yielded no consistent advantage over climatology. Seventeen discrepancies in annual summaries concern fields excluded from the models. Monthly measurements await independent verification, and missing rainy-season observations affect forecast eligibility. Retrospective station selection and sensitivity to the seasonal model limit generalization.
- 18 November 2026
- Published by MDPI AG in Preprints.org
Open AccessAssessing Markov Persistence Beyond Seasonality in Monthly Rainfall-State Prediction in Chad
Does the previous month’s rainfall state improve forecasts once the calendar month is known? We used monthly rainfall records from Chad’s national meteorological agency (ANAM) for 2010–2020. The records cover 60 stations and 7,107 of 7,920 possible station-months. We analyzed four historical stations and extended the comparison to 30 stations selected for completeness, including the original four. An event was a monthly accumulation ≥ 20 mm. We compared calendar-month frequencies, homogeneous and two-season Markov chains, and seasonal logistic regressions with and without the previous state. Across 1,735 eligible rolling-origin forecasts in 2016–2020, pooled Brier scores were 0.05755 for calendar-month climatology, 0.07918 for the two-season Markov chain and 0.05789 for harmonic regression with a lag. The two-season chain outperformed climatology at one station. Adding the lag could improve a given seasonal regression, but the comparison yielded no consistent advantage over climatology. Seventeen discrepancies in annual summaries concern fields excluded from the models. Monthly measurements await independent verification, and missing rainy-season observations affect forecast eligibility. Retrospective station selection and sensitivity to the seasonal model limit generalization.
- Open Access
- 12 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 85207-85216
Hybrid Naïve Bayes Models for Scam Detection: Comparative Insights From Email and Financial Fraud
Online scams continue to escalate in scale and sophistication, ranging from deceptive phishing emails to complex financial fraud schemes. These evolving threats have surpassed the capabilities of traditional detection systems, creating an urgent demand for scalable, real-time, and interpretable machine learning solutions. This study revisits the Naïve Bayes algorithm—often underestimated in modern cybersecurity—as a core building block for effective scam detection. By looking at two specific case studies on phishing emails and financial fraud detection, we look into the math behind Naïve Bayes and talk about its flaws, such as the assumption that features are independent, problems with high-dimensional data, and extreme class imbalance. We apply advanced feature engineering techniques and introduce hybrid model architectures that integrate Naïve Bayes with deep learning and ensemble methods. Our empirical results reveal that a strategically optimized Naïve Bayes model can deliver competitive accuracy, while maintaining transparency and computational efficiency—key attributes for real-world fraud prevention systems. By bridging theory and application, this research challenges the notion that simplicity implies ineffectiveness. It demonstrates that Naïve Bayes, particularly when hybridized, remains a powerful and adaptable tool in the fight against evolving cyber threats, contributing to the design of robust and responsive cybersecurity frameworks.
- 12 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 85207-85216
Open AccessHybrid Naïve Bayes Models for Scam Detection: Comparative Insights From Email and Financial Fraud
Online scams continue to escalate in scale and sophistication, ranging from deceptive phishing emails to complex financial fraud schemes. These evolving threats have surpassed the capabilities of traditional detection systems, creating an urgent demand for scalable, real-time, and interpretable machine learning solutions. This study revisits the Naïve Bayes algorithm—often underestimated in modern cybersecurity—as a core building block for effective scam detection. By looking at two specific case studies on phishing emails and financial fraud detection, we look into the math behind Naïve Bayes and talk about its flaws, such as the assumption that features are independent, problems with high-dimensional data, and extreme class imbalance. We apply advanced feature engineering techniques and introduce hybrid model architectures that integrate Naïve Bayes with deep learning and ensemble methods. Our empirical results reveal that a strategically optimized Naïve Bayes model can deliver competitive accuracy, while maintaining transparency and computational efficiency—key attributes for real-world fraud prevention systems. By bridging theory and application, this research challenges the notion that simplicity implies ineffectiveness. It demonstrates that Naïve Bayes, particularly when hybridized, remains a powerful and adaptable tool in the fight against evolving cyber threats, contributing to the design of robust and responsive cybersecurity frameworks.
- Open Access
- 16 November 2024
- Published by MDPI AG in Machine Learning and Knowledge Extraction
- Vol. 6 (4) , 2639-2658
Application of Bayesian Neural Networks in Healthcare: Three Case Studies
This study aims to explore the efficacy of Bayesian Neural Networks (BNNs) in enhancing predictive modeling for healthcare applications. Advancements in artificial intelligence have significantly improved predictive modeling capabilities, with BNNs offering a probabilistic framework that addresses the inherent uncertainty and variability in healthcare data. This study demonstrates the real-world applicability of BNNs through three key case studies: personalized diabetes treatment, early Alzheimer’s disease detection, and predictive modeling for HbA1c levels. By leveraging the Bayesian approach, these models provide not only enhanced predictive accuracy but also uncertainty quantification, a critical factor in clinical decision making. While the findings are promising, future research should focus on optimizing scalability and integration for real-world applications. This work lays a foundation for future studies, including the development of rating scales based on BNN predictions to improve clinical outcomes.
- 16 November 2024
- Published by MDPI AG in Machine Learning and Knowledge Extraction
- Vol. 6 (4) , 2639-2658
Open AccessApplication of Bayesian Neural Networks in Healthcare: Three Case Studies
This study aims to explore the efficacy of Bayesian Neural Networks (BNNs) in enhancing predictive modeling for healthcare applications. Advancements in artificial intelligence have significantly improved predictive modeling capabilities, with BNNs offering a probabilistic framework that addresses the inherent uncertainty and variability in healthcare data. This study demonstrates the real-world applicability of BNNs through three key case studies: personalized diabetes treatment, early Alzheimer’s disease detection, and predictive modeling for HbA1c levels. By leveraging the Bayesian approach, these models provide not only enhanced predictive accuracy but also uncertainty quantification, a critical factor in clinical decision making. While the findings are promising, future research should focus on optimizing scalability and integration for real-world applications. This work lays a foundation for future studies, including the development of rating scales based on BNN predictions to improve clinical outcomes.
- 28 February 2018
- Published by Pushpa Publishing House in Advances and Applications in Statistics
- Vol. 52 (1) , 1-31
ON DETECTION OF EPILEPTIC SEIZURE WITH AN APPROACH BASED ON POWER SPECTRAL DENSITY WITH AN AR MODEL
- 28 February 2018
- Published by Pushpa Publishing House in Advances and Applications in Statistics
- Vol. 52 (1) , 1-31
ON DETECTION OF EPILEPTIC SEIZURE WITH AN APPROACH BASED ON POWER SPECTRAL DENSITY WITH AN AR MODEL
- 2 June 2015
- Published by Pushpa Publishing House in JP Journal of Biostatistics
- Vol. 12 (1) , 15-31
AUTOMATIC DETECTION OF EPILEPTIC SEIZURE BY EXTRACTING STATISTICALS FEATURES FROM EEG SIGNALS
This paper considers a problem of detection of EEG signals. For this purpose, we propose a criterion based on univariate maximum value of the signal in order to detect the onset of severe possible seizure. We construct a hypothesis test based on this criterion according to the distribution form of our data in order to use the likelihood for the estimation of parameters. Our main contribution is the automatic detection of the breaks of seizure, as the examination of EEG signals is often done via visual inspection of the amplitude in the temporal domain by neurologist practitioners. The results of numerical simulations of the test and application to real EEG data are presented. The accuracy of detection is estimated to 64.88%, with the sensitivity estimated as 83.33%. Received: January 28, 2015 Revised: June 30, 2015 Accepted: July 11, 2015
- 2 June 2015
- Published by Pushpa Publishing House in JP Journal of Biostatistics
- Vol. 12 (1) , 15-31
AUTOMATIC DETECTION OF EPILEPTIC SEIZURE BY EXTRACTING STATISTICALS FEATURES FROM EEG SIGNALS
This paper considers a problem of detection of EEG signals. For this purpose, we propose a criterion based on univariate maximum value of the signal in order to detect the onset of severe possible seizure. We construct a hypothesis test based on this criterion according to the distribution form of our data in order to use the likelihood for the estimation of parameters. Our main contribution is the automatic detection of the breaks of seizure, as the examination of EEG signals is often done via visual inspection of the amplitude in the temporal domain by neurologist practitioners. The results of numerical simulations of the test and application to real EEG data are presented. The accuracy of detection is estimated to 64.88%, with the sensitivity estimated as 83.33%. Received: January 28, 2015 Revised: June 30, 2015 Accepted: July 11, 2015