@inproceedings{scholars18986,
           pages = {249--252},
           title = {Development of Machine Learning Data based Agriculture Monitoring System},
         journal = {2023 IEEE International Conference on Sensors and Nanotechnology, SENNANO 2023},
             doi = {10.1109/SENNANO57767.2023.10352568},
            year = {2023},
            note = {cited By 0; Conference of 2023 IEEE International Conference on Sensors and Nanotechnology, SENNANO 2023 ; Conference Date: 26 September 2023 Through 27 September 2023; Conference Code:195657},
          author = {Iskandar, M. A. F. B. M. and Hawari, H. F. and Kit, K. C. and Ahmad, I.},
        keywords = {Discriminant analysis; Economics; Farms; Forecasting; Internet of things; Learning systems; Machine learning; Neural networks; Soils, Agriculture monitoring; Economic growths; Internet of thing; Learning data; Linear discriminant analyze; Machine learning models; Machine-learning; Monitoring system; Naive bayes; Soil sensors, Crops},
             url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85182737670&doi=10.1109\%2fSENNANO57767.2023.10352568&partnerID=40&md5=cd6b127105d71f5b19683fe3682cc4b6},
        abstract = {Providing food and driving economic growth are vital roles of agriculture. Most farmers still use outdated farming methods that may not keep pace with the country's quick industrialisation and expanding populace, further compounding the problem. This project addresses issues in Malaysia's agriculture sector, such as farmers' reliance on traditional and inefficient farming methods and the limited use of Machine Learning models. In addition, existing approaches consume precious resources since farmers utilise them based on approximations. This project develops a real-time agriculture monitoring system using an Arduino Uno, a soil sensor, and an ESP32 Wi-Fi module. Okra was selected since it can produce crops for 10 to 12 weeks. Okras crops were planted until harvest stage, and the dataset collected from the soil sensor was sent to a ThingSpeak cloud server. Machine Learning models, such as Naive Bayes (NB), Linear Discriminant Analysis (LDA), and Artificial Neural Network (ANN), were used to predict crop health based on the dataset collected. The three selected Machine Learning models performed well in predicting crop health, with NB achieving 99.19 accuracy, LDA achieving 95.82 accuracy, and ANN achieving 94.94 accuracy. It was also found that NB and LDA are recommended for a simpler dataset and prediction, whereas ANN is suitable for more complex use. Overall, this project demonstrates that IoT technologies and Machine Learning can improve agricultural practices by reducing the time and effort required for crop inspection and resource wastage. {\^A}{\copyright} 2023 IEEE.}
}