Abstract : This study employs machine learning techniques to enhance agricultural prediction to secure future food supplies and support sustainable practices. It aims to forecast trends in highland vegetable crops in Benguet providing invaluable insights for enhancing yields and contributing to vital food security initiatives. Notably, the researchers concentrated on crop production data due to inconsistencies and incompleteness in historical datasets collated. Furthermore, it proposes a mobile platform for data collection from farmers, capturing their production and farming practices. This method intends to streamline the data collection process, enabling LGUs to effectively collate and utilize the information. The methodology employed a structured approach, encompassing data collection, preparation, analysis, modeling, and evaluation using Python programming and Power BI tools. In the analytical phase, a variety of machine learning models were explored, including Linear Regression, Lasso Regression, Ridge Regression, Decision Trees, SVM, and Random Forest. Model evaluation was based on the accuracy, MAE, MSE, and RMSE. The Random Forest model emerged as the most suitable choice, boasting the best metrics for production purposes with an accuracy of 98.94%. The outcomes of this study hold significant potential, not only in reshaping agricultural practices and decision-making but also in fostering sustainable approaches to highland vegetable cultivation.

Keyword : Crop Production, Data Mining, Machine Learning, Predictive Analytics, Sustainable Agriculture.