search

stefanonardo / pytorch-esn

An Echo State Network module for PyTorch.
MIT License
211 stars 43 forks source link

Regarding Results Evaluation #1

Closed connectedhealthcaremy closed 6 years ago

connectedhealthcaremy commented 6 years ago

we have run your algorithm and got the following results on the sample model tensor([[ 0.4779], [ 0.2621], [ 0.1902], ..., [ 0.3912], [ 0.3592], [ 0.3210]]) We do not know how to get the classification accuracy please guide us we will be really thankful to you

stefanonardo commented 6 years ago

I'm going to release a new version next week with many improvements and fixes, including the ability to train the network for a classification task.

stefanonardo commented 6 years ago

@connectedhealthcaremy I updated the code (this new version works with PyTorch 0.4 only). On the front page you will notice that I added a parameter for classification tasks, namely output_steps.

Before this update, you had to pass a target value for each timestep of the timeseries, i.e. size had to be (timesteps, batch, feature). Now, if you pass 'mean' or 'last' to output_steps, your target's size must be (1, batch, feature).

In the feature I will add some examples, both for regression and classification.