Spectral estimation from simulations via sketching
This provides a more efficient method for spectral estimation in simulations, such as molecular dynamics, with significant data compression benefits.
The paper tackled the problem of compressing simulation data to estimate time autocorrelation and power spectral density, achieving 90% accuracy with only 10% of the data in a molecular dynamics simulation of methanol.
Sketching is a stochastic dimension reduction method that preserves geometric structures of data and has applications in high-dimensional regression, low rank approximation and graph sparsification. In this work, we show that sketching can be used to compress simulation data and still accurately estimate time autocorrelation and power spectral density. For a given compression ratio, the accuracy is much higher than using previously known methods. In addition to providing theoretical guarantees, we apply sketching to a molecular dynamics simulation of methanol and find that the estimate of spectral density is 90% accurate using only 10% of the data.