TIHT with a relative-error CP approximation is shown to recover bounded-factor low CP-rank tensors under an exponentially strong tensor RIP, whose sample complexity is exponential.
Robust principal component analysis?,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.NA 1years
2019 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Iterative Hard Thresholding for Low CP-rank Tensor Models
TIHT with a relative-error CP approximation is shown to recover bounded-factor low CP-rank tensors under an exponentially strong tensor RIP, whose sample complexity is exponential.