Ensemble Machine Learning Approaches for CPUE Forecasting in Kenya’s Artisanal Marine Fisheries: Application of Xgboost and Prophet with Random Forest Feature Selection

Magak C; Oscar Ngesa; Herbert Imboga

doi:10.51584/IJRIAS.2026.11070114

International Journal of Research and Innovation in Applied Science (IJRIAS)

Ensemble Machine Learning Approaches for CPUE Forecasting in Kenya’s Artisanal Marine Fisheries: Application of Xgboost and Prophet with Random Forest Feature Selection

byMagak C; Oscar Ngesa; Herbert Imboga

Published August 8, 2026 • Vol. 11, Issue 7, pp. 1651–1657Open Access
DOI: 10.51584/IJRIAS.2026.11070114

Download PDF Browse this issue

Abstract

Catch Per Unit Effort (CPUE) remains a fundamental index of fishery productivity, supporting assessments of stock status and guiding management decisions. In many data-poor fisheries, including those in the Western Indian Ocean (WIO), CPUE forecasting is constrained by incomplete time series, high variability, and limited analytical capacity. Recent advances in machine learning (ML) offer powerful alternatives to traditional statistical models by capturing nonlinear dynamics, interactions among variables, and temporal patterns in noisy data.

Keywords: CPUE, XGBoost, Prophet, Random Forest, machine learning

Journal	International Journal of Research and Innovation in Applied Science (IJRIAS)
ISSN	2454-6194
Volume / Issue	Volume 11, Issue 7
Pages	1651–1657
Publication date	August 8, 2026
DOI	10.51584/IJRIAS.2026.11070114
Publisher	RSIS International
License	Open Access

How to cite this article

Magak C, Oscar Ngesa, & Herbert Imboga (2026). Ensemble Machine Learning Approaches for CPUE Forecasting in Kenya’s Artisanal Marine Fisheries: Application of Xgboost and Prophet with Random Forest Feature Selection. International Journal of Research and Innovation in Applied Science (IJRIAS), 11(7), 1651-1657. https://doi.org/10.51584/IJRIAS.2026.11070114

BibTeX

@article{Magak2026,
  title   = {Ensemble Machine Learning Approaches for CPUE Forecasting in Kenya’s Artisanal Marine Fisheries: Application of Xgboost and Prophet with Random Forest Feature Selection},
  author  = {Magak C and Oscar Ngesa and Herbert Imboga},
  journal = {International Journal of Research and Innovation in Applied Science (IJRIAS)},
  volume  = {11},
  number  = {7},
  pages   = {1651--1657},
  year    = {2026},
  doi     = {10.51584/IJRIAS.2026.11070114},
  publisher = {RSIS International}
}