Phoneme {fdWasserstein}

Phoneme data

Description

The dataset comprises 4509 log-periodograms computed from digitalized speech frames. Each log-periodograms is of length 256, and is based on the pronunciation of one of the following five phonemes: "sh", "dcl", "iy", "aa" and "ao".

Usage

data(phoneme)

Format

• logPeriodogram: a 4509x256 matrix containing the log-periodograms.

• Phoneme: a vector of length 4509 containing the phonemes.

Source

The data set was downloaded from the "Elements of statistical learning" website at https://hastie.su.domains/ElemStatLearn/

References

T. Hastie and R. Tibshirani and J. Friedman (2009) The elements of statistical learning: Data mining, inference and prediction, 2nd edn, New York: Springer.

Examples

data(phoneme)
old <- par(mfrow=c(3,2))
for (i in unique(Phoneme)) 
    matplot(t(logPeriodogram[Phoneme==i,]), type="l", 
            xlab="", ylab="", ylim=c(0,30), main=i)
par(old)