Supplementary Material for "Should we sample a time series more frequently? Decision support via multirate spectrum estimation (with discussion)"

TL;DR

This paper discusses a model for estimating the spectral density of stationary time series with missing data, analyzing the cost-benefit of increasing sampling frequency for better spectral estimation.

Contribution

It introduces a multirate spectrum estimation model that accounts for missing data and evaluates the cost implications of sampling rate adjustments.

Findings

Effective spectral estimation with missing data

Cost analysis of sampling frequency changes

Guidance for sampling decision-making

Abstract

This technical report includes an assortment of technical details and extended discussions related to paper "Should we sample a time series more frequently? Decision support via multirate spectrum estimation (with discussion)", which introduces a model for estimating the log-spectral density of a stationary discrete time process given systematically missing data and models the cost implication for changing the sampling rate.