Statistical pattern recognition techniques: A sample problem of ultrasonic determination of interfacial weakness in adhesive joints

Dickstein, P. A.; Segal, Y.; Segal, E.; Sinclair, A. N.

doi:10.1007/bf00566585

Cited by 13 publications

(6 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Inserting values of fi and fi corresponding to points from one side of the dividing curve into Eq. (7) always yields a positive value of dq; consistently negative values of dq are obtained from data points on the opposite side.…”

Section: Deterininntion Of Decision Fruictions For Classijicationmentioning

confidence: 99%

See 1 more Smart Citation

Integrated Statistical Processing of NDE Signals for Class Discrimination

Dickstein

1992

Research in Nondestructive Evaluation

View full text Add to dashboard Cite

An integrated methodology for the analysis of experimental NDE signals for the classification of test-objects according to their mechanical characteristics is presented. The methodology consists of several techniques, adopted from various multidisciplinarial fields, and incorporates concepts from the theory of information, pattern recognition techniques, cluster analysis, and data compression. Statistical tests are performed to verify the significance of the differences between the signals obtained from different groups and to estimate the confidence level associated with the classification procedure. The techniques are outlined and discussed, and experimental examples are presented to demonstrate their successful application.

show abstract

Section: Deterininntion Of Decision Fruictions For Classijicationmentioning

confidence: 99%

“…The task of calculating P, and p e may be eased by determining a mathematical expression for the cluster distributions [7]. For a single production sample Q, with a RAB value denoted by RAE, the probability that it is a member of Class A or Class B can now be calculated from the g functions:…”

Section: (11)mentioning

confidence: 99%

Integrated Statistical Processing of NDE Signals for Class Discrimination

Dickstein

1992

Research in Nondestructive Evaluation

View full text Add to dashboard Cite

show abstract

“…Measuring the deviation of the observed HOC from the expected HOC of Gaussian white noise, is a recommended procedure. A successful application of this idea has been reported in Dickstein et al(1989) in discriminating between ultrasound echo signatures obtained from bonded aluminum specimens that underwent different surface treatment prior to bonding. To measure deviation from white noise, Dickstein et al used the 0' statistic discussed in Kedem and Slud(1982), but using observed rather than expected HOC from white noise.…”

Section: -Yn(o) As a Descriptive-diagnostic Toolmentioning

confidence: 99%

“…Suppose the squared gain function IH(w)f2 is monotone increasing in[C, 7r]. Denote by pi(H) the first-order correlation in the filtered process.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Higher Order Crossings from a Parametric Family of Linear Filters

Kedem¹,

Li²

1989

View full text Add to dashboard Cite

where 8 is the backward shift. In the case of a stationary Gaussian process, DO , FAN., 1473 EDITION OF' I MOV o is OUOLETE S/N 0102-LF-014-6601 SECURITY CLASSIFICATION OF THIS PAGE (ften Date Itntored) SECURITY CLASSFICATION Of THIS PAGE (Whm Dra EaeMO 20. Abstract (continued) the expected zero-crossing rate indexed by a E (-1,1), is always decreasing, completely determines the correlation generating function, and is equal to a constant if and only if the process is a pure sinusoid with probability one. When the process is a sinusoid plus white noise, a sequence {E[D], j = 1,2 .... } is constructed that after proper normalizaj tion converges to the frequency of the sinusoid regardless of the signal to noise ratio. A real data example is given in tracking the vocal sound of a meuaptera novaeangZiae whale.

show abstract