Deep Learning Accelerates Spectral Density Estimation for Functional Time Series
Analysis
Key Takeaways
- •Proposes a deep learning estimator for spectral density of functional time series.
- •Avoids computation of large autocovariance kernels, enabling faster computation.
- •Validated with simulations and application to fMRI images.
“Our estimator can be trained without computing the autocovariance kernels and it can be parallelized to provide the estimates much faster than existing approaches.”