Predicting E-commerce Sales Forecasting and Inventory Management Based on Fuzzy LIM-CNN Technique

Online shopping's share of the market has grown over the last decade, even if retail as a whole has slowed. Logistical costs for EC are higher than for other retail models due to a variety of problems with inventory management. Even though EC makes inventory management more difficult, it works well with EC because consumers can easily compare prices, which is important because goods demand is very price sensitive. Three stages, including preprocessing, feature selection, and model training, make up this suggested approach. The term "data preprocessing" refers to the steps used to remove any anomalies or noise from the chosen data. Using feature selection, we may sift through different datasets and extract the most unnecessary and superfluous information. A Fuzzy LIM-CNN was employed for the purpose of training the model. The proposed approach outperforms CNN and Fuzzy LIM with an average accuracy of 91.67%.