A Hybrid EfficientNetB0–XGBoost Framework for Betel Leaf Health Classification Using Field Images

• Self-Collected Dataset: A real-world Piper betle leaf
disease dataset personally collected under natural field
conditions in Rajshahi, Bangladesh.
• Hybrid Framework: An EfficientNetB0–XGBoost
framework for betel leaf disease classification.
• Feature Selection: XGBoost-based selection of informa-
tive deep features for efficient classification.
• Optimal Representation: Identification of an effective
feature representation for improved classification perfor-
mance.
• Comparative Analysis of ML Models: Comparative
evaluation of multiple machine learning models to iden-
tify the most effective model for disease classification.