Smoothness Prior Approach Research Articles

Smoothness prior approach to explore mean structure in large-scale time series

This article is addressed to the problem of modeling and exploring mean value structure of large-scale time series data and time-space data. A smoothness prior modeling approach (Smoothness Prior Analysis of Time Series, Lecture Notes in Statistics, vol. 116, Springer, New York, 1996.) is taken. In this approach, the observed series are decomposed into several components each of which are expressed by smoothness priors models. In the analysis of POS and GPS data, various useful information were extracted by this decomposition, and result in discoveries in these areas.

An advanced detrending method with application to HRV analysis.

An advanced, simple to use, detrending method to be used before heart rate variability analysis (HRV) is presented. The method is based on smoothness priors approach and operates like a time-varying finite-impulse response high-pass filter. The effect of the detrending on time- and frequency-domain analysis of HRV is studied.

Method to Subtract an Effect of the Geocorona EUV Radiation from the Low Energy Particle(LEP) Data by the Akebono (EXOS-D) Satellite.

It was found that the LEP data was sometimes contaminated with unexpectedly larger noise mainly due to the geocorona EUV radiation. Its noise level is so high that we are obliged to substract the background noise in ground processing when we intend to analyze such LEP data. We therefore propose an efficient and useful procedure for eliminating this background noise, which is highly implemented so as to reduce the computing time. This method is based on a Bayesian smoothness prior approach with a state space modeling.

Non-Gaussian Smoothness Prior Approach to Irregular Time Series Analysis

A non Gaussian smoothness prior approach to the analysis of irregular time series is shown. The model is expressed in state space form with non Gaussian system and/or observational noise. Filtering and smoothing algorithms are realized by numerical representation of related distributions. Significant merit of non Gaussian modeling is demonstrated by some numerical examples.