New independent component analysis tools for time series

Independent component analysis is a popular approach in search of latent variables and structures in high-dimensional data. We propose extensions of classical FOBI and JADE estimates for multivariate time series, with a special focus on time series with stochastic volatility.