Time series classification using LibSVM

It is possible to use LibSVM for time series classification, based on the raw previous values or on some set of features associated to those previous values.

In this case we are using raw values and it is possible to observe that the following parameters are important to get a good performance:

• Sliding Window Length (SWL): here please refer to my previous post
• C = Regularization parameter, as you can see in this figure:

• Gamma = ?defining how far the influence of a single training example reaches, with low values meaning ‘far’ and high values meaning ‘close’, as per this equation

You can test the code here: https://github.com/multiopti/MYWAI/blob/main/ts_svr.ipynb

Feel free to leave your comments here below, I would be happy to answer.

At?MYWAI?we promote agile, explainable, reliable and affordable ML at the edge.