Crop Yield Prediction in Diverse Environmental Conditions
Using Ensemble Learning
Associate Professor, Department of AIML, Institute of Aeronautical Engineering, Hyderabad,
India. Dr.skjakeerhussain@gmail.com
Associate
Professor, Department of CSE, CMR College of Engineering & Technology ,
Hyderabad, India
3*Dr. G. Jose Moses
Professor of CSE, Malla Reddy University, Hyderabad,
India
Mail ID: josemoses@gmail.com
4D. Ramanjaneyulu
Associate Professor, Department of AIML, CMR College of Engineering &
Technology , Hyderabad, India
Assistant Professor, Department of Computer
Science and Engineering, Malla
Reddy University, Maisammaguda, Hyderabad, India
ORCID:- 0009-0005-2774-192X
sreekanthkottu@gmail.com
6Dr K. Vasanth Kumar,
Professor, Department of Computer Science and
Engineering, Malla Reddy (MR) Deemed to be University, Maisammaguda, Dhulapally,
Hyderabad, India vasanthkamatham@gmail.com
*corresponding author:
Abstract
Analysis of crop yield prediction is essential for the effective
decision-making process in agriculture, this is a significant part for food
safety and optimal resource management with the domain of undeniable
environmental fluctuations. This work features a development for forecasting
crop return via an ensemble training approach paired with a bio-inspired
optimization method. A novel ensemble classifier based on adaptive boosting
(AdaBoost) is adopted to boost the forecasting accuracy by combining weak
learners. However, the predictive performance is strongly influenced by the
number of relevant hyperparameters and relevant feature subsets. In order to
resolve this dilemma, we propose a whale-inspired optimization strategy which
can perform feature selection and hyperparameter tuning with equal
effectiveness. Estimation of the planned framework through multi-year
agricultural datasets covering various crops, soil kinds, climates, and areas.
The validated results show that the optimized ensemble model achieves better
accuracy and stability than conventional ML methods. The results imply that the
presented approach is effective in predicting crop yield in heterogeneous environments
and can practically contribute to designing reliable decision-support systems
for sustainable and climate-smart agriculture.
Keywords:
Crop yield forecasting; Ensemble learning; Metaheuristic optimization; Feature
selection; Precision agriculture; Climate-aware prediction
