深圳宝安网站设计公司,甘肃省建设厅门户网站,做哪个网站有效果,高速公路建设网站#x1f4a5;#x1f4a5;#x1f49e;#x1f49e;欢迎来到本博客❤️❤️#x1f4a5;#x1f4a5; #x1f3c6;博主优势#xff1a;#x1f31e;#x1f31e;#x1f31e;博客内容尽量做到思维缜密#xff0c;逻辑清晰#xff0c;为了方便读者。 ⛳️座右铭欢迎来到本博客❤️❤️ 博主优势博客内容尽量做到思维缜密逻辑清晰为了方便读者。 ⛳️座右铭行百里者半于九十。 本文目录如下 目录 1 概述 2 运行结果 3 参考文献 4 Matlab代码实现 1 概述
“本文提出了一种基于字典的L1范数稀疏编码用于时间序列预测不需要训练阶段参数调整最少适用于非平稳和在线预测应用。预测过程被表述为基础追求 L1 范数问题其中为每个测试向量估计一组稀疏权重。尝试了约束稀疏编码公式包括稀疏局部线性嵌入和稀疏最近邻嵌入。16个时间序列数据集用于测试离线时间序列预测方法其中训练数据是固定的。所提出的方法还与Bagging树BT最小二乘支持向量回归LSSVM和正则化自回归模型进行了比较。所提出的稀疏编码预测显示出比使用10倍交叉验证的LSSVM更好的性能并且比正则化AR和Bagging树的性能明显更好。平均而言在LSSVM训练时可以完成几千个稀疏编码预测。
2 运行结果 部分代码
clear all; %Time series Prediction using Sparse coding with overcomplete dictionaries %In each case, the test data prediction is plotted versus the real data %and the sparsity of the solution is recorded. %both L1-magic (if LASSO0) and CVX libraries (if LASSO1) must be included % in the Matlab path
%if normalize1 use sqrt(x*x), normalize2 use st.dev., normalize3 use %the L1 norm, or zero then no normalization. normalize12; normalize22; eps0.001; %the error constraint thr0.001; %the pruning threshold NN20000; %these are the max number of neighbors allowed dthr0.0; %the distance threshold used to filter the dictionary. If it %is zero then no dictionary filtering is done LASSO1; %0 for BP and 1 for BPDN or LASSO using CVX
for kkk1:16 %The 16 data sets used for evaluation nnnnkkk; if(nnnn1) %Mackey-Glass data load MGData; a MGData; time a(:, 1); x_t a(:, 2); trn_data zeros(500, 5); chk_data zeros(500, 5); time 1:sz; Train x_t(1:100); Test x_t(101:190); K6; eps0.001; C USD-EURO Data elseif(nnnn15) load IkedaData1; %Z-normalized if(nonorm1) for i1:L1-K dzz(i)1; end end
end
%Now we normalize the targets of the training data for i1:L1-K if(normalize15) T(i) (trg1(i)-dmm(i))/dvv(i); else T(i) trg1(i)/dzz(i); end end TR T; %%%%%%%%This is the dictionary filtering process (if we want to reduce the %%%%%%%%number of similar atoms. It is controlled by the dthr value %%%%%%%%specified by the user. I have not investigated this a lot dictsizesize(DD); nndictsize(1); %the large dimension mmdictsize(2); %the small dimension RRrandn(nn,mm); RRorth(RR); tooclose0; for io1:nn %over all the atoms xio DD(io,1:mm); cnt0; for jo io1 : nn dddd(jo) dist(xio,DD(jo,1:mm)); if(dddd(jo) dthr) cntcnt1; %one or more atoms are too close end end if(ionn) mindist(io) min(dist(xio,DD(io1:nn,1:mm))); %the min distance for each atom with the next ones else mindist(io)0; end if(cnt0) %no atoms are too close FF(io,1:mm) DD(io,1:mm); FT(io) TR(io); else %some atoms are too close, so we remove this one and put a random atom FF(io,1:mm) RR(io, 1:mm); FT(io) 0; %the target for the random atoms is zero tooclosetooclose1; end end %So, now the new dictionary is FF and the new targets is FT %(if no filtering happened then FF is the same as DD) TooClose(kkk) tooclose; %this will tell us how many atoms were replaced (removed) MinDist(kkk,1:nn)mindist(1:nn); %here we construct the test data Test(1:(L2-K),1:K) tst(1:(L2-K),1:K); TT trg2(1:(L2-K));
ML2-K;
tic %to get the test time sprs0; %to accumulate the sparsity over test vectors
for i1:M %loop over M vectors from the test data disp(**************); disp(i); test(1:K) Test(i,1:K); %here we normalize the test vector by its own dot product if(normalize21) normtest sqrt(test*test); elseif(normalize22) normtest sqrt(var(test)); elseif(normalize23) normtest norm(test,1); elseif(normalize20) normtest1; end
3 参考文献 部分理论来源于网络如有侵权请联系删除。 [1]Waleed Fakhr, Sparse Locally Linear and Neighbor Embedding for Nonlinear Time Series Prediction, ICCES 2015, December 2015.
4 Matlab代码实现
