Skip to content
L-MLDS modeling; Tensor time series prediction
MATLAB M
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data
functions
functions_4D
LICENSE
Nasdaq100_diag.m
Nasdaq100_full.m
README.md
Read me.txt
Video_diag.m
Video_full.m
sst_diag.m
sst_full.m
tesla_diag.m
tesla_full.m
test_4D.m

README.md

L-MLDS-for-Tensor-Time-Series

L-MDS modeling; Tensor time series prediction

The folder named 'data' contains the datasets.

The folder named 'functions' contains all functions we need.

%----------------------------------------------------------

Running the files named-

sst_diag,

sst_full,

Video_diag, 

Video_full,

nasdaq100_diag, 

nasdaq100_full,

Tesla_diag, 

Tesla_full

to verify the results in paper.

All the experiment parameters are set as described in the paper.

%----------------------------------------------------------

Experimental results are shown as prediction errors and prediction values.

The prediction errors are shown in figure 1.

The prediction values are shown in figure 2.

The prediction values for all datasets are saved as:

Result_lds.mat---------by LDS
Result_mlds.mat--------by MLDS
result_dct.mat---------by dct-MLDS
result_dft.mat---------by dft-MLDS
result_dwt.mat---------by dwt-MLDS

for checking.

%-----------------------------------------------------------

How to check a specific value in prediction result?

Ans: If one wants to check a specific value in the output tensor series, please modify the tensor subscript in the following program (which is used to draw the prediction trend in each matlab file described above):

%% Prediction value------------------------------------------------------------

figure(2)

real_lds=zeros(1,Ntest);

real_mlds=zeros(1,Ntest);

real_dct=zeros(1,Ntest);

real_dwt=zeros(1,Ntest);

real_dft=zeros(1,Ntest);

real=zeros(1,Ntest);

for i=1:Ntest

real_lds(i)  = Result_lds{i}(1,2);               % modify the subscript here
real_mlds(i) = Result_mlds{i}(1,2);              % modify the subscript here
real_dct(i)  = result_dct{i}(1,2);               % modify the subscript here
real_dwt(i)  = result_dwt{i}(1,2);               % modify the subscript here
real_dft(i)  = result_dft{i}(1,2);               % modify the subscript here
real(i)      = X{i+Ntrain}(1,2);                 % modify the subscript here

end

%% ------------------------------------------------------------------------------

We provide the 4D L-MLDS model with discrete cosine transform.

Run test_4D.m to check the prediction results.

The results are shown as prediction trends in Figure(1), to check different stocks please do as described above.

You can’t perform that action at this time.