International Journal of Professional Practice

Deep Learning Approach for Detection and Prediction of Pest Infections on Plants in Greenhouses

2025-11-11T10:35:53+00:00

Pest infestations remain problematic in greenhouse agriculture, lowering yields and increasing costs. Manual pest monitoring is laborious, slow, and error-prone, resulting in delayed interventions and excessive pesticide applications. The main objective of this research was to develop an AI-driven hybrid deep learning model for automated pest detection and outbreak prediction, integrating Convolutional Neural Networks (CNNs) for image-based classification and Long Short-Term Memory (LSTM) networks for forecasting to improve response efficiency. This study relied on secondary datasets, such as PlantVillage and IP02, owing to constraints in obtaining primary datasets. These datasets provided pre-annotated pest images and historical climate trends, guaranteeing model robustness. Although the model was trained using secondary data, the study was contextualized by greenhouses in Limuru, Naivasha, and Thika; areas where pest control is an ongoing challenge. To enhance the model's ability to generalize and perform well in an array of agricultural environments, a stratified sampling method which considered farm size and agroclimatic differences was used. Techniques such as image augmentation, noise reduction, and normalization of features were utilized to further improve the quality of the data before the model was trained. Model training and optimization were performed in a GPU-enabled Google Colab environment, which supported batch processing, early stopping, and fine-tuning of hyperparameters. The hybrid model achieved 94.7% accuracy, 93.6% precision, 92.8% recall, and a 93.2% F1-score. With a Mean Absolute Error (MAE) of 0.14 and an R² score of 0.89, the LSTM forecasting module demonstrated its efficiency. This hybrid approach enables early pest identification, preventative actions, and reduced pesticide use.

An Ensembled Tabnet-Based Model Approach for Diabetes Disease Classification

2026-02-10T16:38:02+00:00

Despite the advancements in machine learning (ML) for classification tasks, accurately classifying diseases on limited-feature medical datasets remains challenging. Traditional ML models struggle with interpretability, necessitating an exploration of novel technique. This research developed and evaluated a novel TabNet-based ensemble model for diabetes classification, rating its performance against Extreme Gradient Boosting (XGBoost), Random Forest and base TabNet models. The study utilized the PIMA Indian Diabetes dataset from a public ML Repository, which contains 768 tuples (8 features and 1 outcome variable). A TabNet-based ensemble model was developed using a weighted averaging strategy. For comparative analysis, baseline models, including XGBoost, Random Forest, and a standalone TabNet model were also implemented and optimized. Model performance was assessed using key metrics: balanced accuracy, precision and recall (class 1), F1 score, and Receiver Operating Characteristic-Area Under the Curve (ROC-AUC). The ensembled TabNet-based model consistently achieved the highest performance metrics: balanced accuracy of 83%, precision of 84%, recall of 89%, F1 score of 84%, and ROC-AUC of 90.4% compared to XGBoost (accuracy 81% , precision 79% , recall 86%, F1 score 81%, ROC-AUC 88.6%) , Random Forest (accuracy 81%, precision 78%, recall 87%, F1 score 81%, ROC-AUC 91.6%) and base TabNet (accuracy 81%, precision 80%, recall 82%, F1 score 81%, ROC-AUC 86.7%). The study recommends healthcare institutions to adopt the validated ensemble TabNet-based architecture as a standardized framework for clinical decision support systems across multiple diseases. Further, researchers should establish this methodology as the preferred approach for limited-feature medical datasets, extending beyond diabetes to include cardiovascular, hypertension, and cancer screening applications.