Japanese Vowels

<p>This dataset records 640 time series of 12 LPC cepstrum coefficients taken from nine male speakers. The dataset consists of multivariate time series (12 variables) of length 29 time steps. Each time series is associated with a class (1-9) indicating the speaker. The dataset consists of 270 time series used for training and 370 for testing.</p> <p>This is a pre-processed version saved in numpy format. The original Japanese Vowels dataset and additional information can be found on <a href="https://archive.ics.uci.edu/dataset/128/japanese+vowels">UCI</a>.</p> <p>The data are 3-dimensional arrays of shape [n_samples, time_steps, n_variables]. The data can be loaded as follows:</p> <pre><code>loaded_data = np.load("Japanese_Vowels.npz") Xtr = loaded_data['Xtr'] # Training data of shape (270, 29, 12) Ytr = loaded_data['Ytr'] # Training labels of shape (270, 1) Xte = loaded_data['Xte'] # Test data of shape (370, 29, 12) Yte = loaded_data['Yte'] # Test labels of shape (370, 1)</code></pre>