Modelling of Ladyfinger Plantations Using Open Data Malaysia: A Comparative Study of Machine Learning Techniques

Abstract

This study investigates the potential of using open data from Open Data Malaysia to develop classification models for agricultural practices specifically focusing on ladyfinger plantations. The integration of climate data with agricultural data is performed to build predictive models for crop yield prediction for sustainable agriculture. Four machine learning models, namely Naïve Bayes, SVM, KNN, and decision tree, are evaluated based on various performance metrics. The ladyfinger dataset was obtained from Open Data Malaysia containing both climate and agricultural data. was preprocessed and mined. The results indicate that the Naïve Bayes model achieves the highest performance making it the most suitable model for predicting ladyfinger yield. The decision tree model performed poorly and may not be suitable for this type of classification task. This study highlights two important findings. Firtstly, the inclusion of climate data significantly improved the classification performance of the models. Secondly the limited size of the ladyfinger dataset emphasizes the need for larger and more diverse datasets to enhance the accuracy and generalizability of predictive models in agriculture. Open data initiatives are important for providing researchers with data however larger and more diverse datasets are needed to improve model accuracy. Future research could investigate machine learning models for predicting crop yields in different crops with various climate and agricultural data combinations.