32 isolet 1 **Author**: Ron Cole and Mark Fanty (cole@cse.ogi.edu, fanty@cse.ogi.edu) **Donor**: Tom Dietterich (tgd@cs.orst.edu) **Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/ISOLET) - 1994 **Please cite**: **ISOLET (Isolated Letter Speech Recognition)** This data set was generated as follows. 150 subjects spoke the name of each letter of the alphabet twice. Hence, we have 52 training examples from each speaker. The speakers are grouped into sets of 30 speakers each, 4 groups can serve as trainings set, the last group as the test set. You will note that 3 examples are missing. I believe they were dropped due to difficulties in recording. I believe this is a good domain for a noisy, perceptual task. It is also a very good domain for testing the scaling abilities of algorithms. For example, C4.5 on this domain is slower than backpropagation! Past Usage: * Fanty, M., Cole, R. (1991). Spoken letter recognition. In Lippman, R. P., Moody, J., and Touretzky, D. S. (Eds). Advances in Neural Information Processing Systems 3. San Mateo, CA: Morgan Kaufmann. Goal: Predict which letter-name was spoken, a simple classification task. 95.9% correct classification using the OPT backpropagation implementation. Training on isolet1+2+3+4, testing on isolet5. Network architecture: 56 hidden units, 26 output units (one-per-class). * Dietterich, T. G., Bakiri, G. (1991) Error-correcting output codes: A general method for improving multiclass inductive learning programs. Proceedings of the Ninth National Conference on Artificial Intelligence (AAAI-91), Anaheim, CA: AAAI Press. Goal: same as above. 95.83% correct using OPT backpropagation. (Architecture: 78 hidden units, 26 output units, one-per-class). 96.73% correct using a 30-bit error-correcting output code with OPT (Architecture: 156 hidden units, 30 output units). **Attributes** All attributes are continuous, real-valued attributes scaled into the range -1.0 to 1.0. The features are described in the paper by Cole and Fanty cited above. The features include spectral coefficients; contour features, sonorant features, pre-sonorant features, and post-sonorant features. Exact order of appearance of the features is not known. 1 ARFF 1994 2014-08-20T20:59:05 Public https://test.openml.org/data/v1/download/32/isolet.arff 32 class study_14 public active 2024-01-10 13:50:33 7c00e71486690813d4a60181bcbca7f6