The Role of Artificial Neural Networks in Mathematical Modeling: From Single Architecture to Hybrid Frameworks in Environmental System

2026-04-18

PRE

Sukhpal Singh

Kulwinder Singh Parmar *

Sarbjit Singh

DOI：10.1016/j.asoc.2026.115236

Original

Original request for help

Save

Abstract

En 中文

• Comprehensive Literature Survey: Reviewed over 100 high-impact journal articles on the application of Artificial Neural Networks (ANN) in environmental forecasting. • Fundamentals of ANN: Explained ANN architecture, learning algorithms, activation functions, and their ability to capture non-linear patterns in environmental time series data. • Standalone ANN Applications: Highlighted studies where ANN was applied individually to forecast environmental parameters such as such as air quality, temperature, public health, and rainfall. • Hybrid ANN Models: Explored hybrid modelling approaches combining ANN with methods like ARIMA, wavelet, GA-ANN, PSO-ANN to improve prediction accuracy for complex environmental datasets. • Performance Insights and Trends: Summarized comparative strengths, limitations, and emerging trends in ANN-based environmental modeling, providing guidance for future research.

Keywords:

Artificial Neural Networks

Environmental Forecasting

Hybrid Models

Non-linear Patterns

Mathematical Modeling

The Role of Artificial Neural Networks in Mathematical Modeling: From Single Architecture to Hybrid Frameworks in Environmental System

Abstract

Journal

Organization