Klasifikasi Penjualan berdasarkan Platform pada UMKM Omah Branded Menggunakan Random Forest

Abstract

UMKM have a role in development growth to increase state income. Omah Branded which is an UMKM in the Fashion category from small children to adults in the Mungkid area, Magelang Regency, Indonesia. Currently the resulting sales transaction data has not been used to classify or classify sales products based on the sales platform that can affect the revenue of the Branded Omah. Selling products on online platforms is often referred to as digital marketing. It has become widespread and widely applied in Indonesia due to the development of the internet and changing consumers. Easy internet access using wifi or gadgets makes it easier for people to access information about a product or service they are looking for. One of the data mining for classification is the Random Forest Algorithm. The Random Forest algorithm has a random selection in generating child nodes for each node (top node), the classification of each tree is accumulated and the classification results that appear frequently can improve accuracy. In this study, by classifying Omah Branded sales data based on the sales platform using the random forest method, it is hoped that the results of this study can be used as a development solution for taking Omah Branded marketing strategies. The accuracy value using the Random Forest classification on the sales data of this Omah Branded product produces an accuracy of 92% based on the results of the confusion matrix calculation.