Supply chain resilience and supplier reliability are crucial for highly perishable agricultural goods. The disruptions in these supply chains can severely affect perishable goods, leading to revenue loss. This research proposes an approach that integrates iterative simulation and machine learning techniques to assess the supply chain's resiliency and predict the supplier's reliability in the perishable strawberry supply chain. The methodology begins by generating data for involved entities to replicate real-world dynamics. A discrete-event simulation environment helps in modeling intricate supply chain interactions, capturing time-dependent events, and evaluating the impact of disruptions in a controlled virtual environment. The model forecasts supplier reliability scores, introducing dynamism and uncertainty into the simulation. The model is trained for machine learning, training, and evaluating multiple classifiers, such as XGBoost, Random Forest, SVM, Logistic Regression, KNN, Naive Bayes, and Decision Tree, for predictive and comparative performance. Sensitivity analysis is conducted to test the robustness of the predictive model by varying input parameters. Integrating this predictive model with the perishable product supply chain enhances decision-making, enabling proactive measures to be taken in response to potential supplier shutdowns. The proposed approach provides a comprehensive and practical solution for improving supply chain resilience and managing supplier reliability in the strawberry supply chain. Integrating simulation and machine learning techniques helps the supply chain decision-makers optimize involved operations and mitigate disruptions, benefiting the entire food supply chain ecosystem. Hence, this research offers valuable insights into improving the supply chain for perishable goods and supplier reliability. The findings help decision-makers devise strategies and timely choices to sustain a robust and reliable supply chain.