Agriculture information system for horticulture based on machine learning.

Agriculture is one of the important sectors in Indonesia. There are several commodities consumed by Indonesian citizens, one of them is shallots as a cooking spice and herbal medicine. Several problems occasionally arise in the shallots cultivation process; one of them is increasing plant pest organism's attacks. However, the control process is still done conventionally. Therefore, it is necessary to build a system to diagnose diseases automatically. This paper proposes an information system for agriculture shallots based on machine learning using dempster Shafer. After collecting the data, dempster Shafer is applied to determine and define a reasonable level of confidence and a function to evaluate a possibility. Based on the experiment, the accuracyofmax belief for detection of shallots disease is 60%. By using several scenarios, the experimental result shows that dempster Shafer can detect shallot disease well. Furthermore, the algorithm can be implemented in web and mobile applications based to monitor the controlling process easier. [ABSTRACT FROM AUTHOR]