Decision Support System for Land Selection to Increase Crops Productivity in Jember Regency Use Learning Vector Quantization (LVQ)

Abstract

Indonesia is one of the countries that can produce a wide variety of agricultural products. Most of Indonesia’s population depend on agriculture because the land to produce food crops is quite fertile and productive. However, as the population grows, the use of converted agricultural land creates difficulties for its own expansion. Aside from inadequate agricultural land conditions, poor agricultural yields caused by prolonged seasons and bad weather are some of the problems faced in Indonesia, especially in Java. The focus of this study is to apply Learning Vector Quantization (LVQ), which is part of the artificial neural network method, to provide recommendations from three types of plants that are most suitable for planting based on land or regional conditions. The recommendations generated by using the Learning Vector Quantization (LVQ) method show several significant differences when compared to real conditions, that is equal to 64.51% compared to 35.49%. While the comparison of LVQ method recommendations compared with expert recommendations shows a percentage of 93.54% compared to 6.46%. This shows that in reality many plants are still planted based on low land suitability.