Archives of Current Research International

The present study develops a nonlinear Neural Network Autoregressive model with Exogenous inputs (NNARX) to model the Gross Cropped Area (GCA) in Kerala, India, using historical annual data from 1965-1966 to 2022-2023. The primary objective is to assess the predictive performance of NNARX models relative to benchmark Neural Network Autoregressive (NNAR) models by incorporating key climatic variables such as annual rainfall, minimum temperature, and maximum temperature as exogenous inputs. Several model configurations were examined, and the NNARX (2, 2) structure, which employs two lagged values of both endogenous and exogenous inputs, was identified as the optimal model based upon standard model selection criteria such as Akaike Information Criteria (AIC), Corrected AIC (AICc) and Bayesian Information Criteria (BIC) values. Model performance was evaluated using Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE), with results showing that the NNARX (2, 2) model significantly outperformed the NNAR alternatives. The performance of the NNARX model underscores the importance of incorporating climatic factors to more accurately capture the nonlinear and dynamic relationships that influence agricultural land use. The findings highlight that climate variables play a crucial role in determining the cropping patterns and GCA in Kerala. Moreover, the results demonstrate that machine learning based time series approaches, such as NNARX, offer powerful tools for agricultural modelling, capable of providing more accurate and reliable forecasts. By offering a data-driven framework to predict production outcomes, assess the trade-offs of various management strategies, and create policies that support economic benefits, food security, and environmental sustainability, particularly in light of growing climate variability, these models can support evidence-based agricultural planning, effective resource allocation, and the development of sustainable development strategies.