How to apply Dynamic Time Warping (dtw) to a continuous time series signal with multiple templates and no start or endpoints.

Hello,

I am looking to apply dynamic time warping with a continous time series signal. I have multiple templates that i would like the DTW to provide the most suitable match to.

I understand this may be difficult as there is no start or end points in the time series data.

Does anyone have any examples of a simular case study or any suggestions to get me started?

Thanks!

Answer 1

[Sara AlaeeJordehi]

Hello,

I have a paper that finds the DTW motifs in long and continuous time series. Here is the link:

https://arxiv.org/abs/2009.07907

Let me know if you need any additional information.

Best,

Sara

[Thomas Wellsbury]

Hello Sara,

Many thanks for offering your support and providing your paper, looks promising.

Can i ask if you have a MATLAB exaple of the Scalable Warping Aware Matrix Profile in code?

Thanks,

Tom

[Sara AlaeeJordehi]

Hi Thomas,

No problem. Yes. I do have the Matlab source codes. It is on the supporting webpage of the paper. You can find it on the following link, under Codes section:

https://sites.google.com/site/dtwmotifdiscovery/

Please let me know if you needed any additional information.

Best,

Sara

[Thomas Wellsbury]

Hi Sara,

Many thanks once again for getting back to me.

The aim of the codes look very simular to my project which is very promising.

I have sent an email for futher infomation, please let me know if you have not recieved the email.

Tom