Nazwa Arraudhah

IMPROVING SALES CLASSIFICATION OF FASHION PRODUCTS AT SABHIRA OFFICIAL WITH RANDOM FORES

Introduction

Improving sales classification of fashion products at sabhira official with random fores. Improve fashion product sales classification at Sabhira Official using Random Forest. This research achieved 99.81% accuracy, providing insights for stock and promotion strategies.

Abstract

Abstract This research focuses on improving the accuracy of the fashion product sales classification model at Sabhira Official Store by applying the Random Forest algorithm. The approach used follows the stages of Knowledge Discovery in Database (KDD), which includes data selection, preprocessing, transformation, data mining, and evaluation. The research data consisted of 1,559 transactions in the period August to October 2023, with attributes such as product category, number of items sold, price, and sales category (low, medium, high). The model was developed using RapidMiner software, with a data split of 70% for training and 30% for testing. The analysis results show that the Random Forest algorithm is able to achieve an accuracy rate of 99.81%, with precision for the “High” category reaching 100%, while other categories have values above 99%. Evaluation using confusion matrix shows a very low prediction error rate, so this model can classify sales levels more accurately. The results of this study provide useful insights for Sabhira Official Store in stock management and data-driven promotion strategies. Keywords— Random Forest, sales classification, fashion products, KDD.

Review

This paper, titled "IMPROVING SALES CLASSIFICATION OF FASHION PRODUCTS AT SABHIRA OFFICIAL WITH RANDOM FORES," presents a focused study on enhancing sales prediction accuracy for a specific e-commerce fashion retailer. The research aims to develop a robust model for classifying fashion product sales into 'low,' 'medium,' or 'high' categories, providing valuable insights for operational decisions. Utilizing the well-established Knowledge Discovery in Database (KDD) methodology, the authors systematically approach the problem from data selection to evaluation, ensuring a structured and replicable process. This targeted application of data mining techniques promises tangible benefits for Sabhira Official Store's inventory and marketing strategies. A significant strength of this study lies in its clear methodology and exceptionally strong empirical results. The application of the Random Forest algorithm to a dataset of 1,559 transactions from August to October 2023 yielded an impressive overall accuracy rate of 99.81%. Furthermore, the model demonstrates high precision across all sales categories, with the ‘High’ category achieving a perfect 100% and others exceeding 99%. The use of RapidMiner for model development and a standard 70/30 data split for training and testing adds to the credibility of the findings. The reported very low prediction error rate, as evidenced by the confusion matrix, strongly supports the model's reliability in accurately classifying sales levels. While the reported accuracy is outstanding and highly beneficial for Sabhira Official, the review could suggest avenues for broader impact and deeper understanding. Future work might involve exploring the model's generalizability by applying it to different fashion retailers or expanding the dataset to cover longer periods, incorporating seasonal trends. Additionally, an analysis of feature importance within the Random Forest model could offer deeper business intelligence, revealing which specific attributes (e.g., product category, price points) are most influential in determining sales levels. This would not only enhance the model's predictive power but also provide actionable insights beyond mere classification, further optimizing stock management and promotion strategies in a dynamic fashion market.

Full Text

Comments

You need to be logged in to post a comment.