isolet

**Author**: Ron Cole and Mark Fanty ([e-mail], [e-mail])
**Donor**: Tom Dietterich ([e-mail])
**Source**: [UCI](https://archive.ics.uci.edu/ml/datasets/ISOLET)
**Please cite**: UCI

### Description

ISOLET (Isolated Letter Speech Recognition) dataset was generated as follows: 150 subjects spoke the name of each letter of the alphabet twice. Hence, there are 52 training examples from each speaker. The speakers are grouped into sets of 30 speakers each, 4 groups can serve as training set, the last group as the test set. A total of 3 examples are missing, the authors dropped them due to difficulties in recording.

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!

### Source

* Creators:
Ron Cole and Mark Fanty
Department of Computer Science and Engineering,
Oregon Graduate Institute, Beaverton, OR 97006.
cole '@' cse.ogi.edu, fanty '@' cse.ogi.edu

* Donor:
Tom Dietterich
Department of Computer Science
Oregon State University, Corvallis, OR 97331
tgd '@' cs.orst.edu

### Attributes Information

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 below.
The features include spectral coefficients; contour features, sonorant features, pre-sonorant features, and post-sonorant features. The exact order of appearance of the features is not known.

### Relevant papers

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.

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.

Dietterich, T. G., Bakiri, G. (1994) Solving Multiclass Learning Problems via Error-Correcting Output Codes.

**作者**：罗恩·科尔（Ron Cole）与马克·凡蒂（Mark Fanty）（邮箱：[e-mail]、[e-mail]）
**数据捐赠者**：汤姆·迪特里希（Tom Dietterich）（邮箱：[e-mail]）
**数据来源**：[UCI机器学习仓库](https://archive.ics.uci.edu/ml/datasets/ISOLET)
**引用要求**：请引用UCI相关文献

### 数据集描述
ISOLET（孤立字母语音识别）数据集的生成流程如下：共有150名受试者，每名受试者需将字母表中每个字母的名称朗读两遍，因此每名受试者可提供52条训练样本。所有受试者按每30人一组划分为4组训练集与1组测试集。本次数据集共缺失3条样本，原作者因录制故障将其剔除。
该数据集是噪声感知分类任务的优质研究域，同时也是测试算法缩放能力的绝佳实验平台。例如，在该数据集上，C4.5算法的运行速度慢于反向传播（backpropagation）算法！

### 数据来源说明
* 数据集创建者：
罗恩·科尔、马克·凡蒂
俄勒冈研究生院计算机科学与工程系，比弗顿，俄勒冈州97006
邮箱：cole '@' cse.ogi.edu、fanty '@' cse.ogi.edu
* 数据捐赠者：
汤姆·迪特里希
俄勒冈州立大学计算机科学系，科瓦利斯，俄勒冈州97331
邮箱：tgd '@' cs.orst.edu

### 属性信息
所有属性均为连续实值属性，取值范围已归一化至-1.0至1.0之间。特征的具体定义可参考科尔与凡蒂的相关论文（见下文引用文献）。
特征涵盖频谱系数、轮廓特征、浊音特征、前浊音特征与后浊音特征。特征的准确出现顺序尚不明确。

### 相关参考文献
1. Fanty, M., Cole, R. (1991). 语音字母识别[Spoken letter recognition]. 收录于Lippman, R. P., Moody, J., 与 Touretzky, D. S. 主编. 《神经信息处理系统进展3》（Advances in Neural Information Processing Systems 3）. 加利福尼亚州圣马特奥：摩根·考夫曼出版社（Morgan Kaufmann）.
2. Dietterich, T. G., Bakiri, G. (1991) 纠错输出码：一种改进多分类归纳学习程序的通用方法[Error-correcting output codes: A general method for improving multiclass inductive learning programs]. 收录于第九届全国人工智能会议（AAAI-91）论文集，加利福尼亚州阿纳海姆：美国人工智能协会出版社（AAAI Press）.
3. Dietterich, T. G., Bakiri, G. (1994) 基于纠错输出码的多分类学习问题求解[Solving Multiclass Learning Problems via Error-Correcting Output Codes].