On learning algorithm selection for classification
On learning algorithm selection for classification
复制标题
DOI：
10.1016/j.asoc.2004.12.002
复制
发表时间：
2006-01-01
影响因子：
8.7
通讯作者：
Smith, KA
中科院分区：
文献类型：
--
作者：
Ali, S;Smith, KA
本文介绍了一种新的学习算法评价和选择方法，并给出了基于分类的经验结果。对8种算法/分类器进行了100个不同分类问题的实证研究。我们评估算法的性能方面的各种准确性和复杂性的措施。与“没有免费的午餐”定理一致，我们并不期望确定在所有数据集上表现最好的单一算法。相反，我们的目标是确定数据集的特性，这些数据集通过某些学习算法可以进行上级建模。我们的经验结果被用来生成规则，使用基于规则的学习算法C5.0，来描述哪些类型的算法适合于解决哪些类型的分类问题。大多数规则都是以高置信度等级生成的。(C)2005年爱思唯尔B。V.保留所有权利。
This paper introduces a new method for learning algorithm evaluation and selection, with empirical results based on classification. The empirical study has been conducted among 8 algorithms/classifiers with 100 different classification problems. We evaluate the algorithms' performance in terms of a variety of accuracy and complexity measures. Consistent with the No Free Lunch theorem, we do not expect to identify the single algorithm that performs best on all datasets. Rather, we aim to determine the characteristics of datasets that lend themselves to superior modelling by certain learning algorithms. Our empirical results are used to generate rules, using the rule-based learning algorithm C5.0, to describe which types of algorithms are suited to solving which types of classification problems. Most of the rules are generated with a high confidence rating. (C) 2005 Elsevier B. V. All rights reserved.