本文選題：混沌時間序列預(yù)測 + 回聲狀態(tài)網(wǎng)絡(luò)��； 參考：《工程科學(xué)與技術(shù)》2017年06期

【摘要】：回聲狀態(tài)網(wǎng)絡(luò)(ESN)是一種重要的時間序列預(yù)測方法,但在訓(xùn)練數(shù)據(jù)存在噪聲或野點情況下,ESN將會出現(xiàn)過擬合問題。針對該問題,提出基于平滑消邊絕對偏離罰函數(shù)的回聲狀態(tài)網(wǎng)絡(luò)(SCAD-ESN)模型。不同于在模型中加入嶺回歸、L1范數(shù)罰函數(shù)及小波降噪等常規(guī)方法,該模型利用SCAD罰函數(shù)對變量進(jìn)行選擇,將小變量置為零以滿足變量稀疏性,將大變量直接置為常數(shù),從而能夠很好地解決ESN過擬合問題并滿足近似無偏估計。對于SCAD罰函數(shù)的非凸函數(shù)優(yōu)化問題,提出基于局部二次近似(LQA)的求解方法,將最小角回歸(LQR)方法用于SCAD罰函數(shù)求解,避免了計算量巨大的問題。使用基于粒子群優(yōu)化(PSO)的超參數(shù)選取方法快速確定平滑消邊絕對偏離 回聲狀態(tài)網(wǎng)絡(luò)模型的超參數(shù),克服利用經(jīng)驗選取超參數(shù)時存在的盲目性較大且難以確定整體最優(yōu)的超參數(shù)問題�；煦缦到y(tǒng)數(shù)值仿真和網(wǎng)絡(luò)流量仿真結(jié)果表明,相對于常規(guī)模型,該模型能有效地降低測試誤差,從而克服過擬合問題。
[Abstract]:Echo state network (ESNN) is an important time series prediction method, but if there is noise or outliers in the training data, the ESNN will be overfitted. To solve this problem, a SCAD-ESN-based echo state network model based on smooth edge-elimination and absolute deviation penalty function is proposed. Different from the conventional methods such as ridge regression L1 norm penalty function and wavelet denoising, the model uses SCAD penalty function to select variables, sets small variables to zero to satisfy the sparsity of variables, and places large variables directly as constants. Thus, the problem of ESN overfitting can be solved well and the approximate unbiased estimation can be satisfied. For the non-convex function optimization problem of SCAD penalty function, a method based on local quadratic approximation (LQA) is proposed. The minimum angle regression method is used to solve the SCAD penalty function, which avoids the problem of huge computation. Based on particle swarm optimization (PSO), the super-parameter selection method is used to quickly determine the superparameters of the smooth edge-free absolute deviation from the echo state network model. In order to overcome the problem of blind and difficult to determine the global optimal hyperparameter when using experience to select superparameters. The simulation results of chaotic system and network traffic show that compared with the conventional model, the proposed model can effectively reduce the test error and overcome the problem of over-fitting.
【作者單位】： 河南科技大學(xué)網(wǎng)絡(luò)與通信技術(shù)研究所;河南科技大學(xué)網(wǎng)絡(luò)信息中心;河南科技大學(xué)數(shù)學(xué)與統(tǒng)計學(xué)院;
【基金】：國家自然科學(xué)基金資助項目(11501067) 賽爾網(wǎng)絡(luò)下一代互聯(lián)網(wǎng)技術(shù)創(chuàng)新項目資助(NGII20150508)
【分類號】：TP18

【相似文獻(xiàn)】

相關(guān)期刊論文 前6條

1 李恬鑒;張秉鐸;凌培亮;;罰函數(shù)參數(shù)優(yōu)化設(shè)計法及其在可控硅多環(huán)調(diào)速系統(tǒng)動態(tài)設(shè)計中的應(yīng)用[J];電氣傳動;1985年06期

2 蔡偉建;李津蓉;;一種基于罰函數(shù)的數(shù)據(jù)協(xié)調(diào)模型[J];傳感技術(shù)學(xué)報;2007年01期

3 郭一楠;程健;楊梅;;支持向量回歸超參數(shù)的混沌文化優(yōu)化選擇方法[J];控制與決策;2010年04期

4 楊斌,聶在平;一種支持向量回歸中超參數(shù)自適應(yīng)方法[J];廣西師范大學(xué)學(xué)報(自然科學(xué)版);2003年01期

5 金添;周智敏;宋千;常文革;;基于證據(jù)框架的模糊超球面支持向量機(jī)超參數(shù)優(yōu)化[J];應(yīng)用科學(xué)學(xué)報;2007年03期

6 宋杰;;一種新的無參數(shù)支持向量機(jī)分類器[J];計算機(jī)工程;2007年23期

相關(guān)會議論文 前1條

1 李潔;高峰;管曉宏;周佃民;;支持向量回歸學(xué)習(xí)方法中基于穩(wěn)定型GA的超參數(shù)選擇[A];第二十四屆中國控制會議論文集（下冊）[C];2005年

，

本文編號：1887635