Development of an Enhanced Predictive Model for Road Accident Occurrence in Nigeria
DOI:
https://doi.org/10.55123/jomlai.v5i1.7700Kata Kunci:
Artificial Neural Network, Road Accident Prediction, Nigeria, Machine Learning, Traffic Safety, FRSC, Predictive ModelingAbstrak
Road accidents in Nigeria rank as the second highest globally, with 33.7% of deaths per 100,000 persons occurring annually. This study developed and tested a predictive model for road accident occurrence using Artificial Neural Networks (ANN) to address the technological gap in Nigeria's road safety management systems. A feed-forward neural network architecture comprising 52 input neurons, three hidden layers (32, 16, and 8 neurons) with ReLU activation, and a single sigmoid output neuron was designed. Dropout (0.3, 0.3, 0.2) and L2 regularization (0.001, 0.001, 0.0005) were incorporated to address sample size constraints. The dataset comprised 2,847 records from FRSC, NEMA, and NBS (2018-2023) across twelve Nigerian states, with 24 features spanning road, environmental, driver, and vehicle factors. Stratified random splitting yielded 1,994 training, 570 validation, and 283 temporally distinct test records. The model achieved 84.5% accuracy (95% CI: 79.8%-88.5%), 77.0% recall, 89.4% specificity, and 0.89 AUC on independent test data—a 13.5 percentage point improvement over the existing K-modes system (p<0.0001). Five-fold cross-validation confirmed stability (84.3%±0.6%). Feature importance analysis identified speeding (18.4%), alcohol impairment (15.2%), wet roads (11.8%), night driving (9.4%), and lane discipline (8.1%) as dominant predictors, with human factors accounting for 45.3% of predictive power. This study provides the first evidence-validated ANN-based accident prediction model calibrated for Nigeria, establishing a reproducible methodological template for developing contextually-adapted predictive systems in data-constrained environments while demonstrating statistically significant and practically meaningful improvement over existing approaches.
Referensi
[1] K. Chen, V. Singh, and P. Singh, “Tools in mining software repositories,” in Proc. 13th Int. Conf. Computational Science and Its Applications, IEEE Press, 2021, pp. 89–98.
[2] World Health Organization (WHO), “Analysis of road traffic accidents in modern cities of the world,” Int. Article on Health, Safety and Environment, 2021.
[3] S. Peled, K. Mohammed, and O. Fatma, “Exploiting coarse-grained reuse-based opportunities in big data multi-query optimization,” J. Computational Science, vol. 28, pp. 432–452, 2022.
[4] S. Kumar and D. Tripathi, “A data mining framework to analyze road accident data,” J. Big Data, vol. 2, no. 26, pp. 1–18, 2021.
[5] C. Jaspan, M. Jorde, A. Knight, C. Sadowski, and E. K. Smith, “Advantages and disadvantages of a monolithic repository: A case study at Google,” in Proc. ACM/IEEE 40th Int. Conf. Software Engineering: Software Engineering in Practice (ICSE-SEIP), Gothenburg, Sweden, 2022, pp. 225–234.
[6] M. Shapiro and H. Weatherspoon, “Cloudifying source code repositories: How much does it cost?” ACM SIGOPS Operating Systems Review, vol. 45, no. 2, pp. 1–6, Apr. 2021, doi: 10.1145/1773912.1773919.
[7] C. Maddison and D. Tarlow, “Structured generative models of natural source code,” in Proc. 31st Int. Conf. Machine Learning (ICML), Beijing, China, 2021, vol. 32, pp. 1–14, arXiv:1401.0514v2 [cs.PL].
[8] R. Malhotra and A. Chug, “Software maintainability: Systematic literature review and current trends,” Int. J. Software Engineering and Knowledge Engineering, vol. 26, no. 8, pp. 1221–1253, 2021, doi: 10.1142/S021819406500431.
[9] J. Xu, H. Li, and S. Zhou, “An overview of deep generative models,” IETE Technical Review, vol. 32, no. 2, pp. 131–139, 2021, doi: 10.1080/02564602.2021.987328.
[10] D. A. Arnott and G. Pervan, “Design science in decision support systems research: An assessment using the Hevner, March, Park, and Ram guidelines,” J. Assoc. Information Systems, vol. 13, no. 11, pp. 923–949, 2021.
[11] S. Agarwal and A. Patel, “A study on graph storage database of NoSQL,” Int. J. Soft Computing, Artificial Intelligence and Applications (IJSCAI), vol. 5, no. 1, pp. 33–40, 2021.
[12] A. Dey, “Machine learning algorithms: A review,” Int. J. Computer Science and Information Technologies, vol. 7, no. 3, pp. 1174–1179, 2021.
[13] S. O. Olatunji, S. U. Idrees, Y. S. Al-Ghamdi, and J. S. A. Al-Ghamdi, “Mining software repositories: A comparative analysis,” Int. J. Computer Science and Network Security (IJCSNS), vol. 10, no. 8, pp. 161–174, 2021.
[14] B. Sanchez-Lengeling and A. Aspuru-Guzik, “Inverse molecular design using machine learning: Generative models for matter engineering,” Science, vol. 361, no. 6400, pp. 360–365, 2022.
[15] H. An, D. Jung, and H.-L. Choi, “Deep generative models-based anomaly detection for spacecraft control systems,” Sensors, vol. 23, no. 1991, 2023, doi: 10.3390/s20211991.
Unduhan
Diterbitkan
Cara Mengutip
Terbitan
Bagian
Lisensi
Hak Cipta (c) 2026 Chukwudi Ugbaja, Friday E. Onuodu, Henry Onyebuchukwu Ordu, Emmanuel J. Izionworu
Artikel ini berlisensiCreative Commons Attribution-ShareAlike 4.0 International License.
Copyright (c) 2022 The authors. Published by Yayasan Literasi Indonesia 
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License
The author(s) whose article is published in the JOMLAI journal attain the copyright for their article and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. By submitting the manuscript to JOMLAI, the author(s) agree with this policy. No special document approval is required.
The author(s) guarantee that:their article is original, written by the mentioned author(s),
- has never been published before,
- does not contain statements that violate the law, and
- does not violate the rights of others, is subject to copyright held exclusively by the author(s), and is free from the rights of third parties, and that the necessary written permission to quote from other sources has been obtained by the author(s).
The author(s) retain all rights to the published work, such as (but not limited to) the following rights:
- Copyright and other proprietary rights related to the article, such as patents,
- The right to use the substance of the article in its own future works, including lectures and books,
- The right to reproduce the article for its own purposes,
- The right to archive all versions of the article in any repository, and
- The right to enter into separate additional contractual arrangements for the non-exclusive distribution of published versions of the article (for example, posting them to institutional repositories or publishing them in a book), acknowledging its initial publication in this journal (JOMLAI: Journal of Machine Learning and Artificial Intelligence).
Suppose the article was prepared jointly by more than one author. Each author submitting the manuscript warrants that all co-authors have given their permission to agree to copyright and license notices (agreements) on their behalf and notify co-authors of the terms of this policy. JOMLAI will not be held responsible for anything that may arise because of the writer's internal dispute. JOMLAI will only communicate with correspondence authors.
Authors should also understand that their articles (and any additional files, including data sets, and analysis/computation data) will become publicly available once published. The license of published articles (and additional data) will be governed by a Creative Commons Attribution-ShareAlike 4.0 International License. JOMLAI allows users to copy, distribute, display and perform work under license. Users need to attribute the author(s) and JOMLAI to distribute works in journals and other publication media. Unless otherwise stated, the author(s) is a public entity as soon as the article is published
