Doan Aneta
Semi-supervised learning is a powerful machine learning paradigm that leverages both labeled and unlabeled data to improve model performance. This approach is particularly useful in scenarios where labeled data is scarce or expensive to obtain. Among the various techniques in semi-supervised learning, label propagation has emerged as an effective method for inferring labels for unlabelled instances based on the structure of the data. When formulated on a bipartite graph, closed-form label propagation provides a mathematically efficient way to distribute label information across the dataset, making it a compelling approach for many real-world applications. A bipartite graph is a special type of graph where nodes are divided into two disjoint sets, with edges connecting only nodes from different sets. This structure is commonly found in many domains, such as recommendation systems, bioinformatics, and document classification. By leveraging the inherent relationships within a bipartite graph, label propagation can efficiently distribute label information from labeled to unlabeled nodes. The closed-form solution to label propagation further enhances this method by providing a computationally efficient way to compute labels without requiring iterative updates, which are common in traditional label propagation algorithms.
PDFShare this article
Telecommunications System & Management received 109 citations as per Google Scholar report
