Predicting Consumer Demand Based on Retail Stock Using the K-Nearest Neighbors Algorithm

Inefficient stock management, such as improper stock management, will result in excess or shortage of goods. Excess stock can cause high storage costs and the risk of unsold goods. Predict consumer needs based on stock. Analyze inefficient stock to improve shortages. One effective method for making this prediction is using the K-Nearest Neighbors (K-NN) algorithm. The K-NN algorithm is a simple but powerful machine-learning technique that can be used for classification and regression. The model scenario results show 24 objects in the Low-needs group and 14 in the High-needs group. Evaluation and performance testing using the Rapid Miner tool can also produce a relevant picture of the modelled scenario. The model implemented using the K-NN algorithm has an Accuracy value of 97.50% with a Standard Deviation of +/- 750%, then a Precision value of 100%, and a Recall value of 950%. By measuring model performance with cross-validation, the resulting accuracy has a standard deviation value, which aims to see the distance between the average accuracy and the accuracy of each experiment (iteration)