Analisa Prediksi Kesejahteraan Masyarakat Nelayan Lombok Timur Menggunakan Algoritma Random Forest

Abstract

The economic life of the people on the coast, especially fishermen, is very dependent on the natural resources that are around, for example, marine resources, which still get the top position in the survival of fishing communities which are widely used and are also included as renewable natural resources. One example used as material for this research is the fishing community in East Lombok, West Nusa Tenggara. The Fishermen's Community can be interpreted as a group of people whose main livelihood is fishermen. The characteristics of the life of this community are different from society in general. Natural factors influence their lives a lot, from their lifestyle to the level of their economy and welfare, which is different from other communities. The purpose of this study is to predict the level of welfare of fishing communities in East Lombok, West Nusa Tenggara by using the classification method and the Random Forest algorithm. The dataset used is private data, the data is taken from fishing applications. Data processing is done to get the result or performance of the algorithm as the best result in predicting. From the existing dataset we use five supporting variables including, Education, family members, wells (related to clean water), employment and housing. The results or targets of this data processing are the level of welfare of fishing communities with prosperous and non-prosperous statuses. The final results of this study are seen using the Confusion Matrix, where the end result is the accuracy value. Random Forest has the highest accuracy value with a value of 93.37% and an AUC value of 0.735%.