Crop Yield Prediction Using Hybrid Machine Learning Approach: A Case Study of Lentil (Lens culinaris Medik.)
Problem
Predicting lentil grain yield, a multifactorial nonlinear phenomenon, from numeric agronomic traits.
Why it matters
Accurate yield forecasting supports crop management and breeding decisions for food production.
What the AI did
Hybrid frameworks combining MARS feature selection with SVR and ANN models (MARS-ANN, MARS-SVR) to map predictors to yield.
Results
MARS-based hybrids outperformed individual MARS, SVR and ANN models on RMSE, MAD, MAPE using 518 lentil genotypes; exact values not stated
Impact
Hybrid models beat individual baselines on the lentil dataset; no field deployment stated in evidence.
Record
| Venue | Agriculture |
|---|---|
| Published | 2023 |
| Work type | article |
| Authors | Pankaj Das, Girish Kumar Jha, Achal Lama, Rajender Parsad |
| Citations | 40 (OpenAlex, at import time) |
| References | 49 |
| Model or method | MARS-ANN and MARS-SVR hybrid models |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3390/agriculture13030596 |
| Open access | gold |
| Keywords | crop yield, MARS, SVR, ANN, hybrid machine learning, lentil |
Importance rating
incremental / niche
Lentil yield forecasting supports crop management and breeding, but abstract only demonstrates model superiority on 518 genotypes; no deployment or broader food-production scale stated.