Info on Phoneme Data Set
This data set was obtained from www-stat.stanford.edu/ElemStatLearn and was used as one of example sets in Hastie et al.: The elements of statistical learning. Following is the data provided with this data set.
These data arose from a collaboration between Andreas Buja, Werner Stuetzle and Martin Maechler, and we used as an illustration in the paper on Penalized Discriminant Analysis by Hastie, Buja and Tibshirani (1995), referenced in the text. The data were extracted from the TIMIT database (TIMIT Acoustic-Phonetic Continuous Speech Corpus, NTIS, US Dept of Commerce) which is a widely used resource for research in speech recognition. A dataset was formed by selecting five phonemes for classification based on digitized speech from this database. The phonemes are transcribed as follows: "sh" as in "she", "dcl" as in "dark", "iy" as the vowel in "she", "aa" as the vowel in "dark", and "ao" as the first vowel in "water". From continuous speech of 50 male speakers, 4509 speech frames of 32 msec duration were selected, approximately 2 examples of each phoneme from each speaker. Each speech frame is represented by 512 samples at a 16kHz sampling rate, and each frame represents one of the above five phonemes. The breakdown of the 4509 speech frames into phoneme frequencies is as follows: aa ao dcl iy sh 695 1022 757 1163 872 From each speech frame, we computed a log-periodogram, which is one of several widely used methods for casting speech data in a form suitable for speech recognition. Thus the data used in what follows consist of 4509 log-periodograms of length 256, with known class (phoneme) memberships. The data contain 256 columns labelled "x.1" - "x.256", a response column labelled "g", and a column labelled "speaker" identifying the diffferent speakers.
