| 引用本文: | 韩敏,穆大芸.基于Adaboost算法的回声状态网络预报器[J].控制理论与应用,2011,28(4):601~604.[点击复制] |
| HAN Min,MU Da-yun.Improvement of echo state network accuracy with Adaboost[J].Control Theory and Technology,2011,28(4):601~604.[点击复制] |
|
| 基于Adaboost算法的回声状态网络预报器 |
| Improvement of echo state network accuracy with Adaboost |
| 摘要点击 2571 全文点击 2131 投稿时间：2010-01-18 修订日期：2010-05-04 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/j.issn.1000-8152.2011.4.CCTA100060 |
| 2011,28(4):601-604 |
| 中文关键词 ESN Adaboost. RT算法 非线性时间序列 预测 |
| 英文关键词 ESN Adaboost. RT algorithm nonlinear time series prediction |
| 基金项目 国家自然科学基金资助项目(60674073); 国家高技术研究发展“863”计划资助项目(2007AA04Z158); 国家科技支撑计划资助项目(2006BAB14B05). |
|
| 中文摘要 |
| 把单个回声状态网络(echo state network, ESN)的预测模型作改进, 对整体ESN预测精度的提高是有限的. 针对以上问题, 本文考虑整体ESN. 首先利用Adaboost算法提升单个ESN的泛化性能及预测精度, 并且根据Adaboost算法的结果, 建立一种ESN预报器(Adaboost ESN, ABESN). 这个ESN预报器根据拟合误差不断修正训练样本的权重, 拟合误差越大, 训练样本权重值就越大; 因此, 它在下一次迭代时, 就会侧重在难以学习的样本. 把单个ESN的预测模型经过加权, 然后按照加法组合在一起, 形成最终的ESN预测模型. 将该预测模型应用于太阳黑子、Mackey-Glass时间序列的预测研究, 仿真结果表明所提出的预测模型在实际时间序列预测领域的有效性. |
| 英文摘要 |
| Modifying the prediction model of individual echo state network(ESN) improves the total prediction result with limited extent. To solve this problem, we consider an ensemble of ESN. The general performance and prediction accuracy of each individual ESN is boosted by using the Adaboost algorithm. Based on the Adaboost algorithm results, we develop an ESN predictor(ABESN). In this predictor, the weights of training samples are constantly adjusted according to the fitting error, the greater the fitting error, the heavier the weights for the training samples. Therefore, the ESN predictor will focus on the hard-learning samples in the next iteration cycle. The prediction models of individual ESN are weighted and added up to form the final predictor of the ensemble of ESN. The presented model is tested on the benchmark prediction problem of Mackey-Glass time series as well as the time series of sunspots. Simulation results demonstrate its high prediction accuracy and effectiveness. |
|
|
|
|
|