Abstract:

This dataset contains MATLAB-based coded algorithms for k-means, self-organizing maps, and spectral clustering of echolocation clicks. It was used to analyze recorded audio files from bottom-moored passive acoustic buoys. The results of these algorithms are presented in R4.x261.000:0014.

Suggested Citation:

Juliette Ioup, Jack LeBien. 2018. Detection and classification algorithms using k-means, self-organizing maps, and spectral clustering to identify beaked whale echolocation clicks from passive acoustic data. Distributed by: GRIIDC, Harte Research Institute, Texas A&M University–Corpus Christi. doi:10.7266/N7W094CG

Publications:

LeBien, J. G., & Ioup, J. W. (2018). Species-level classification of beaked whale echolocation signals detected in the northern Gulf of Mexico. The Journal of the Acoustical Society of America, 144(1), 387–396. doi:10.1121/1.5047435

Purpose:

These detection and classification algorithms were created to analyze passive acoustic recordings collected from bottom-moored EARS buoys in the northern Gulf of Mexico to identify spectral qualities of species-level beaked whale echolocation clicks.

Data Parameters and Units:

Amplitude (volts), click ID (sequential number of click detections), source filename (filename format is as follows: 5dddFFFF.BBD 5 = year 2015, ddd = three-digit yearday, FFFF = hexadecimal index number, BB = buoy number, 01 to 10, D = disk number, 0 to 3), species (BWG = beaked whale of the Gulf), 10dB bandwidth (Hz), 20 dB central frequency (Hz), spectral centroid (Hz), 10 dB duration (microseconds), 95% energy duration (microseconds), 95% amplitude duration (microseconds), 95% Teager-Kaiser duration (microseconds), higuchi fractal dimension, castiglioni fractal dimension, adapted box fractal dimension, Katz fractal dimension, Shannon entropy (bits), Renyi entropy (alpha =3, bits), SEM (standard error of the mean), envelope (volts; calculated envelope of the raw signal), time (seconds), click 1 amplitude (volts; raw signal), click 2 amplitude (volts; generated signal created combining the raw signal with random noise), click 3 amplitude (volts; generated signal created by combining the raw signal with random noise), peak frequency (Hz), mean SNR (dB; average signal to noise ratio for each trial) purity (0 to 1; represents the purity of clustering using each of the methods (SOM - self-organizing map, k-means, and spectral clustering), entropy (represents the entropy of clustering using each of the methods (SOM -self-organizing map, k-means, and spectral clustering), Features (training feature sets used for feedforward neural network classification; HFD -Higucki fractal dimension, WPD energy map - wavelet packet decomposition energy map), classification accuracy (percent)

Individual Files: