Prediction update algorithms for XCSF: RLS, Kalman filter, and gain adaptation
Prediction update algorithms for XCSF: RLS, Kalman filter, and gain adaptation
复制标题
XCSF 的预测更新算法：RLS、卡尔曼滤波器和增益自适应
DOI：
--
复制
发表时间：
2006
期刊：
影响因子：
--
通讯作者：
D. Goldberg
中科院分区：
文献类型：
--
作者：
P. Lanzi;D. Loiacono;Stewart W. Wilson;D. Goldberg
我们研究不同的预测更新算法如何影响 XCSF 的性能。我们考虑三种经典参数估计算法（NLMS、RLS 和卡尔曼滤波器）和四种增益自适应算法（K1、K2、IDBD 和 IDD）。后者已被证明具有与最佳算法（RLS 和卡尔曼）相当的性能，但复杂性较低。我们应用这些算法来更新 XCSF 中的分类器预测，并比较 XCSF 的七个版本在一组实际函数上的性能。我们的结果表明，最著名的算法仍然表现最佳：带有 RLS 的 XCSF 和带有卡尔曼的 XCSF 的性能明显优于其他算法。相比之下，当添加到 XCSF 时，增益自适应算法的性能与 NLMS（最简单的估计算法）相当，与原始 XCSF 中使用的算法相同。然而，执行相似的算法概括起来却不同。例如：具有卡尔曼滤波器的 XCSF 比具有 RLS 的 XCSF 发展出更紧凑的解决方案，并且增益自适应算法比 NLMS 具有更好的泛化能力。
We study how different prediction update algorithms influence the performance of XCSF. We consider three classical parameter estimation algorithms (NLMS, RLS, and Kalman filter) and four gain adaptation algorithms (K1, K2, IDBD, and IDD). The latter have been shown to perform comparably to the best algorithms (RLS and Kalman), but they have a lower complexity. We apply these algorithms to update classifier prediction in XCSF and compare the performances of the seven versions of XCSF on a set of real functions. Our results show that the best known algorithms still perform best: XCSF with RLS and XCSF with Kalman perform significantly better than the others. In contrast, when added to XCSF, gain adaptation algorithms perform comparably to NLMS, the simplest estimation algorithm, the same used in the original XCSF. Nevertheless, algorithms that perform similarly generalize differently. For instance: XCSF with Kalman filter evolves more compact solutions than XCSF with RLS and gain adaptation algorithms allow better generalization than NLMS.