<![CDATA[Electricity is a crucial resource in everyday life, and rising household energy demand requires smarter monitoring and management approaches. Analyzing consumption data enables the discovery of typical energy usage behaviors that support efficient resource planning. Clustering techniques are widely used to group usage profiles without predefined categories, with K-Means being one of the most popular methods because of its speed and practical implementation. However, this algorithm is highly dependent on the initial centroid selection and may generate inaccurate grouping results if trapped in local optima. To overcome these drawbacks, this research combines K-Means with the Greylag Goose Optimization (GGO) algorithm, a nature-inspired metaheuristic that simulates the adaptive navigation and social coordination of migratory grey geese. By enhancing both exploration and exploitation, GGO improves the accuracy of centroid placement and overall clustering performance. The research utilized Individual Household Electric Power Consumption dataset, which consists of minute-by-minute measurements of several electrical attributes. After preprocessing and exploratory analysis, clustering was executed using three approaches: conventional K-Means, GGO, and a hybrid K-Means–GGO model. Based on the Silhouette Score evaluation, clustering performance improved significantly from 0.6236 with standard K-Means to 0.9675 using the hybrid approach. The resulting segmentation provides deeper insights into household consumption behaviors.]]>
