The goal of this repository is to share original datasets create by me; they are useful for training, setup and test of machine learning algorithms and neural networks.
synthetic-words folder contains audio files that reproduce the pronunciation of 20 English words got using synthetic voices available on the Mac via the command say. The audio files are in wav format and are sampled at 22050 Hz mono; for each word there are 325 wav files obtained with 25 different voices performed at different speeds. Please visit here for details about this dataset.
functions folder contains a collection of .csv and .arff files synthetically generated applying analytic functions to a closed subset of their own domain. Please visit here for details about this collection of datasets.
timeseries folder contains a collection of .csv and .arff files synthetically generated applying analytic functions to the values of a sequence representing time. Please visit here for details about this collection of datasets.