Derives explicit formula for causally conditioned directed information rate of Gaussian sequences based on optimal prediction and proves O(N^{-1/2} log N) high-probability error bound for the resulting estimator.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IT 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Non-Asymptotic Error Bounds for Causally Conditioned Directed Information Rates of Gaussian Sequences
Derives explicit formula for causally conditioned directed information rate of Gaussian sequences based on optimal prediction and proves O(N^{-1/2} log N) high-probability error bound for the resulting estimator.