Generalized Canonical Correlation Analysis for Classification
arXiv:1304.7981 · doi:10.1016/j.jmva.2014.05.011
Abstract
For multiple multivariate data sets, we derive conditions under which Generalized Canonical Correlation Analysis (GCCA) improves classification performance of the projected datasets, compared to standard Canonical Correlation Analysis (CCA) using only two data sets. We illustrate our theoretical results with simulations and a real data experiment.
28 pages, 3 figures, 7 tables
References in corpus (1)
Cited by in corpus (10)
- Deep Multimodal Image-Repurposing Detection
- The Labeled Multiple Canonical Correlation Analysis for Information Fusion
- Discovering and Deciphering Relationships Across Disparate Data Modalities
- Manifold Matching using Shortest-Path Distance and Joint Neighborhood Selection
- Synergistic Graph Fusion via Encoder Embedding
- On the Incommensurability Phenomenon
- Encoder Embedding for General Graph and Node Classification
- Sparse GCA and Thresholded Gradient Descent
- Refined Graph Encoder Embedding via Self-Training and Latent Community Recovery
- Non-Euclidean Analysis of Joint Variations in Multi-Object Shapes