Speaker
Description
Adewale, O. A.
FEDERAL COLLEGE OF EDUCATION (TECHNICAL) ASABA DELTA STATE
abdulhakimowolabi@gmail.com
Track: AI, Emerging Technologies and Future Learning
Existing research on dietary phytoconstituents often treats compound detection, classification, and bioactivity assessment as separate tasks, limiting reproducibility, interpretability, and systematic prioritization for tumor prevention studies. This study aims to develop and evaluate an explainable machine learning framework for detecting, classifying, and prioritizing dietary phytoconstituents with tumor preventive potential through the integration of phytochemical, molecular descriptor, physicochemical, and bioactivity data. A computational experimental design will be carried out using publicly available datasets curated from dietary and medicinal plant sources based on predefined inclusion criteria such as compound identity, chemical structure or SMILES notation, availability of descriptors, phytochemical class and relevant bioactivity indicators. Data preparation will include duplicate removal, missing value treatment, molecular descriptor generation, feature engineering and dataset partitioning. Supervised machine learning models including Random Forest, Support Vector Machine, Gradient Boosting, XGBoost, Artificial Neural Networks and Graph Neural Networks will be trained and compared. Model performance will be assessed using accuracy, precision, recall, F1-score, ROC-AUC, confusion matrix analysis and cross-validation. Explainable AI methods particularly SHAP and LIME will be used to identify influential molecular descriptors, phytochemical properties and bioactivity features. The anticipated outcome is an interpretable and reproducible AI framework to assist in phytoconstituent screening, prediction and prioritization for subsequent experimental studies