Deep Dynamic Probabilistic Canonical Correlation Analysis
This provides a versatile probabilistic tool for analyzing sequential datasets, such as in finance, but it appears incremental as it builds on existing probabilistic CCA methods.
The paper tackled the problem of analyzing nonlinear dynamical systems by developing D2PCCA, which integrates deep learning with probabilistic modeling to capture nonlinear latent dynamics, and experimental validation on financial datasets showed its effectiveness.
This paper presents Deep Dynamic Probabilistic Canonical Correlation Analysis (D2PCCA), a model that integrates deep learning with probabilistic modeling to analyze nonlinear dynamical systems. Building on the probabilistic extensions of Canonical Correlation Analysis (CCA), D2PCCA captures nonlinear latent dynamics and supports enhancements such as KL annealing for improved convergence and normalizing flows for a more flexible posterior approximation. D2PCCA naturally extends to multiple observed variables, making it a versatile tool for encoding prior knowledge about sequential datasets and providing a probabilistic understanding of the system's dynamics. Experimental validation on real financial datasets demonstrates the effectiveness of D2PCCA and its extensions in capturing latent dynamics.