Presentation1 in datamining using techn.pptx
- 2. Link analysis Link analysis is a technique used in clustering and classification to identify relationships between data points based on connections, associations, or similarities. It is commonly applied in network analysis, fraud detection, recommendation systems, and bioinformatics. Graph is widely used in modelling advanced structures and patterns.
- 3. Graph-Based Clustering Nodes represent entities, and edges represent relationships. Clusters are formed by finding densely connected subgraphs. Link-Based Similarity Measures: Methods like SimRank or PageRank assess similarity between objects based on their shared connections. Community Detection: Algorithms like Louvain method, Girvan-Newman, and Label Propagation identify naturally occurring communities within a network.
- 4. Link Analysis in Classification In classification, link analysis helps assign labels to data points based on their relationships with labeled examples. Some methods include: Relational Classification :Uses network structure to infer labels (e.g., if many friends of a person have a disease, they might be at higher risk). Semi-Supervised Learning: A few labeled nodes help classify the entire network using algorithms like Label Propagation. Graph Neural Networks (GNNs): Deep learning models that classify nodes by aggregating information from their neighbors.
- 5. Link Analysis in Clustering In clustering, link analysis helps group similar data points based on their relationships or structural connections. Some common approaches include: