Unsupervised Clustering

Unsupervised clustering aims to group unlabeled data points into meaningful clusters based on inherent similarities, revealing hidden structures and patterns without prior knowledge of group assignments. Current research emphasizes improving scalability for large datasets, often employing algorithms like k-medoids enhanced with metaheuristics (e.g., whale optimization) or deep learning approaches such as autoencoders and graph convolutional networks for feature extraction and clustering. These advancements are impacting diverse fields, from medical image analysis (e.g., Alzheimer's subtyping, dermatological pattern discovery) and materials science to power system anomaly detection and financial data analysis, enabling efficient knowledge discovery and improved decision-making in various applications.