Scholarlink

<p>This study presents a novel comparative framework for forecasting economic growth in major oil-exporting African countries by integrating Artificial Neural Networks (ANN) with six nature-inspired optimization algorithms—Particle Swarm Optimization (PSO), Invasive Weed Optimization (IWO), Firefly Algorithm (FA), Artificial Bee Colony (ABC), Cultural Algorithm (CA), and a newly developed COVID-19-based Optimization (COVID). The research introduces ANN-COVID, a novel hybrid model that mimics epidemiological dynamics to optimize learning parameters in neural networks, marking a unique contribution to the literature on economic forecasting. The study further contributes by applying these hybrid models to an extensive macroeconomic dataset covering Angola, Congo, Gabon, and Nigeria from 2000 to 2021, using economic indicators such as GDP, oil prices, oil export, oil revenue, exchange rate, interest rate, inflation, and employment. Empirical findings reveal that hybrid models like ANN-FA, ANN-IWO, and ANN-PSO significantly outperform both standalone ANN and other hybrid models, reducing Root Mean Squared Error (RMSE) by up to 97% in countries such as Angola and Nigeria. These results underscore the potential of optimization-driven ANN models in enhancing the accuracy of economic forecasts in resource-dependent economies. Additionally, the comparative analysis reveals why certain algorithms perform better under varying economic conditions, offering valuable insights for policy-oriented modeling in volatile oil-based economies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Conjugation of Artificial Neural Networks with Nature-Inspired Optimization Algorithms for Predicting the Economic Growth of the Top Oil-Producing Countries in Africa

  • 期刊论文

  • 发表日期：

    2025年10月7日

查看全文
  • Bello Sani Yahaya,
  • Sagiru Mati,
  • Demet Beton Kalmaz,
  • Isah Wada

摘要

This study presents a novel comparative framework for forecasting economic growth in major oil-exporting African countries by integrating Artificial Neural Networks (ANN) with six nature-inspired optimization algorithms—Particle Swarm Optimization (PSO), Invasive Weed Optimization (IWO), Firefly Algorithm (FA), Artificial Bee Colony (ABC), Cultural Algorithm (CA), and a newly developed COVID-19-based Optimization (COVID). The research introduces ANN-COVID, a novel hybrid model that mimics epidemiological dynamics to optimize learning parameters in neural networks, marking a unique contribution to the literature on economic forecasting. The study further contributes by applying these hybrid models to an extensive macroeconomic dataset covering Angola, Congo, Gabon, and Nigeria from 2000 to 2021, using economic indicators such as GDP, oil prices, oil export, oil revenue, exchange rate, interest rate, inflation, and employment. Empirical findings reveal that hybrid models like ANN-FA, ANN-IWO, and ANN-PSO significantly outperform both standalone ANN and other hybrid models, reducing Root Mean Squared Error (RMSE) by up to 97% in countries such as Angola and Nigeria. These results underscore the potential of optimization-driven ANN models in enhancing the accuracy of economic forecasts in resource-dependent economies. Additionally, the comparative analysis reveals why certain algorithms perform better under varying economic conditions, offering valuable insights for policy-oriented modeling in volatile oil-based economies.