Analysis Of Stochastic Signals Research Articles

Timely ICU Outcome Prediction Utilizing Stochastic Signal Analysis and Machine Learning Techniques with Readily Available Vital Sign Data.

The ICU is a specialized hospital department that offers critical care to patients at high risk. The massive burden of ICU-requiring care requires accurate and timely ICU outcome predictions for alleviating the economic and healthcare burdens imposed by critical care needs. Existing research faces challenges such as feature extraction difficulties, low accuracy, and resource-intensive features. Some studies have explored deep learning models that utilize raw clinical inputs. However, these models are considered non-interpretable black boxes, which prevents their wide application. The objective of the study is to develop a new method using stochastic signal analysis and machine learning techniques to effectively extract features with strong predictive power from ICU patients' real-time time series of vital signs for accurate and timely ICU outcome prediction. The results show the proposed method extracted meaningful features and outperforms baseline methods, including APACHE IV (AUC = 0.750), deep learning-based models (AUC = 0.732, 0.712, 0.698, 0.722), and statistical feature classification methods (AUC = 0.765) by a large margin (AUC = 0.869). The proposed method has clinical, management, and administrative implications since it enables healthcare professionals to identify deviations from prognostications timely and accurately and, therefore, to conduct proper interventions.

Sampling Reconstruction of Stochastic Signals– The Roots in the Fifties

In this article we are interested in the beginnings and the development of the sampling theory in signal analysis of stochastic signals, locating these in the early fifties. Besides the most important papers by Parzen (1956), Balakrishnan (1957), Belyaev (1959), and Lloyd (1959) we expose and report on few other interesting articles not widely known, giving an overview of the topic.

A novel approach for analog fault diagnosis based on stochastic signal analysis and improved GHMM

This paper presents a novel fault diagnosis method to diagnose ordinary soft fault and incipient soft fault for linear analog circuit. Due to the presence of tolerance, the parameter variation of analog component obeys a nearly normal distribution, and the linear circuit responses are considered as a stochastic process. Therefore, three kinds of stochastic series are extracted as fault features. Then, fault feature samples are divided into train samples and test samples. The relevant statistic values of them are calculated as final test samples in the stage of fault testing. In order to diagnose stochastic series effectively, an improved GHMM classifier is designed based on the fault occurrence characteristics of analog component. The stochastic sequences are thought as the observation symbols for the proposed GHMM. Simulation circuits and actual circuits are tested to prove the effectiveness of the new method.

An Information-Based Approach to Change-Point Analysis with Applications to Biophysics and Cell Biology

This article describes the application of a change-point algorithm to the analysis of stochastic signals in biological systems whose underlying state dynamics consist of transitions between discrete states. Applications of this analysis include molecular-motor stepping, fluorophore bleaching, electrophysiology, particle and cell tracking, detection of copy number variation by sequencing, tethered-particle motion, etc. We present a unified approach to the analysis of processes whose noise can be modeled by Gaussian, Wiener, or Ornstein-Uhlenbeck processes. To fit the model, we exploit explicit, closed-form algebraic expressions for maximum-likelihood estimators of model parameters and estimated information loss of the generalized noise model, which can be computed extremely efficiently. We implement change-point detection using the frequentist information criterion (which, to our knowledge, is a new information criterion). The frequentist information criterion specifies a single, information-based statistical test that is free from ad hoc parameters and requires no prior probability distribution. We demonstrate this information-based approach in the analysis of simulated and experimental tethered-particle-motion data.

Damage monitoring of reinforced concrete frames under seismic loading using cement-based piezoelectric sensor

Damage process monitoring of concrete structures using acoustic emission (AE) technology has been drawn more and more attention due to its powerful capability. In this paper, a brand new cement-based piezoelectric composite sensor was introduced with improved performance, particularly concerning its AE signals detection capacity in the high frequency domain. Such sensors were embedded into the foundation of reinforced concrete (RC) frames during construction, and appointed to monitor concrete damage due to shake table excitation. A standard ground motion record of 1999 Taiwan earthquake with 840 gal and 1300 gal peak ground accelerations were employed in the tests as the excitation inputs. The signals detected by the sensors were stored and analyzed by commercial available data acquisition devices. Advanced stochastic signal analysis methods were adopted to effectively interpret the frequency domain components and identify the useful information representing the damage processes of the RC frames. The results were compared with the eye observations of structural damage and corresponding cyclic loading tests results. It is shown that the health monitoring method using cement-based piezoelectric composite sensors and advanced stochastic signal analysis are capable of detecting and evaluating the damage process of RC frames due to seismic loading. An effective damage indicator of the RC frames is possible to be evaluated from extracted AE information.

Use of the wavelet method for analyzing electrochemical noise data

AbstractThis research paper shows the application of the mathematical tool “wavelet's” to be used in the interpretation and analysis of electrochemical noise signals. The Fast Fourier Transformation (FFT) is one of the most commonly used techniques for analyzing electrochemical noise. But the application of this methodology to the analysis of stochastic signals does not allow their interpretation in a suitable way. The wavelet's methodology carries out the analysis by means of the construction of finite base functions. In several studies, it has been found out that the function that is closest to the electrochemical oscillations is the one called “coiflets,” which presents a similar geometry to what it is observed when an ionic–electronic interchange occurs during an electrochemical interphase. The analysis that appears here was implemented using the toolbox “LabView™.”

Array analyses of seismic noise at Mt. Vesuvius Volcano, Italy

The properties of the seimic noise wavefield at Mt. Vesuvius Volcano, Italy, are investigated using a dense array of short-period seismometers deployed on the south-west flank of the volcano and operated on 1997–1998. The array had an aperture of 500 m and average station spacing of 40 m. The main goal of this deployment was to investigate the possible presence of correlated signals, which could be attributed to subsurface magmatic and/or hydrothermal activity (volcanic tremor). The spectral properties of the noise wavefield are investigated using data from the arrays and from the summit stations of the permanent seismic network operated by the Vesuvius Observatory. The seismic noise is characterized by three main spectral peaks at frequencies <1, 1–1.5, and 3–4 Hz, which are sustained over the different periods of observations. Measurements of slowness and azimuth indicate that the low-frequency (<1 Hz) signals propagate from the S–SW direction, and are attributed to the marine microseismic noise. At higher frequencies (around 1.2 Hz) the wavefield is characterized by waves whose azimuth of propagation changes with time, overimposed to a sustained signal, which propagates from the east direction. No significant preferential directions of propagation are observed at higher frequencies. The dispersive properties of the wavefield are investigated using the correlation technique first proposed by Aki for the analysis of stochastic signals. The good fit we obtain indicate that the 1–10 Hz noise wavefield is dominated by surface waves. The dispersion characteristics of Rayleigh waves are then inverted for a shallow velocity structure beneath the array, which is in agreement with previous results from both large- and small-scale seismic prospections. Comparison between the dispersion curve for Rayleigh waves and the apparent velocities obtained from frequency-slowness analyses, allows to attribute the analyzed wavefield to the action of shallow sources. The shallow nature of the noise wavefield is also confirmed by polarization analyses, which depict a dominance of horizontal components in the particle motion ellipsoid. This evidence implies that the seismic noise cannot be related to the action of the magmatic reservoir, whose depth has been inferred to be at least 6 km below sea level. The daily changes of spectral power show that the noise signal in the 1–4 Hz frequency band has maximum energy during hours of intense human activity. It is therefore reasonable to attribute the noise origin in this particular frequency band to artificial sources related to the densely urbanized areas surrounding Mt. Vesuvius. Notwithstanding that, the concomitant action of additional natural sources cannot be rejected a priori.

Correlation-spectral transformation of signals in adaptive sequential basis

A procedure for deriving a system of stochastic sequential basis functions is described. It is shown that the use of these functions for correlation and spectral analysis of stochastic signals substantially reduces and simlifies computing operations. A simple method of calculation of the spectrum of variances is given.

Distinguishing between low-dimensional dynamics and randomness in measured time series

The success of current attempts to distinguish between low-dimensional chaos and random behavior in a time series of observations is considered. First we discuss stationary stochastic processes which produce finite numerical estimates of the correlation dimension and K 2 entropy under naive application of correlation integral methods. We then consider several straightforward tests to evaluate whether correlation integral methods reflect the global geometry or the local fractal structure of the trajectory. This determines whether the methods are applicable to a given series; if they are we evaluate the significance of a particular result, for example, by considering the results of the analysis of stochastic signals with statistical properties similar to those of observed series. From the examples considered, it is clear that the correlation integral should not be used in isolation, but as one of a collection of tools to distinguish chaos from stochasticity.

Stochastic signal analysis: Numerical representation and applications to system theory

Stochastic signal analysis: Numerical representation and applications to system theory

Diagnostics of gas—liquid flow patterns in chemical engineering systems

A review of the diagnostical methods of the flow patterns which largely determine the performance of gas—liquid systems encountered widely in chemical engineering practice is presented. The need for objective and reliable diagnostical criteria has been confirmed by the analysis of published studies; confidence in the flow pattern maps used in design and operation is limited. The methods of stochastic signal analysis as applied to the fluctuations of some structural (voidage) and energetic (pressure) parameters are found to be promising. The increasing availability of suitable hardware facilitates on-line diagnostics aimed at the optimization and safety control of industrial equipment.