Optimizing K-Nearest Neighbor Values Using The Elbow Method

Optimizing the K value in the K-Nearest Neighbor (KNN) algorithm is a critical step in enhancing model performance, particularly for tasks related to classification and prediction. The Elbow method has emerged as a widely used technique for determining the optimal K value by analyzing the point where adding more neighbors does not significantly reduce error or improve model performance. This study focuses on the application of the Elbow method for optimizing the K value in the context of classifying BRIMo application reviews on the Google Play platform. By conducting a comparative analysis between the K value determined by the Elbow method and the traditionally chosen K value in KNN, this research demonstrates the advantages of using the Elbow method for improving classification accuracy. The experimental results reveal that the K value derived from the Elbow method (0.8) outperforms the traditional K value (0.78), resulting in a notable improvement in model performance. This superior accuracy highlights the effectiveness of the Elbow method in selecting a more optimal K value, leading to a better fit for the data and more precise predictions. Consequently, the use of the Elbow method for optimizing the K value in the KNN algorithm proves to be a highly effective strategy for improving classification models, especially when applied to real-world data such as application reviews on digital platforms like Google Play.