lucapedrelli / deepesn Goto Github PK

Deep Echo State Network (DeepESN) Python library

License: BSD 3-Clause "New" or "Revised" License

Python 100.00%

deepesn's Introduction

Deep Echo State Network Python library (DeepESNpy), 2018

Deep Echo State Network (DeepESN) is an extension of the ESN model towards the deep learning paradigm (Deep Reservoir Computing). A DeepESN is a deep Recurrent Neural Network composed by a hierarchy of recurrent layers intrinsically able to develop hierarchical and distributed temporal features. Such characteristics make DeepESN suitable for time-series and sequences processing.

All details about DeepESN model are described in the reference paper (CITATION REQUEST):
C. Gallicchio, A. Micheli, L. Pedrelli,
"Deep Reservoir Computing: A Critical Experimental Analysis", Neurocomputing, 2017, vol. 268, pp. 87-99

The design of DeepESN model in multivariate time-series prediction tasks is described in the following paper:
C. Gallicchio, A. Micheli, L. Pedrelli,
"Design of deep echo state networks", Neural Networks, 2018, vol. 108, pp. 33-47

AUTHOR INFORMATION

Luca Pedrelli
[email protected]
[email protected]

Department of Computer Science - University of Pisa (Italy)
Computational Intelligence & Machine Learning (CIML) Group
http://www.di.unipi.it/groups/ciml/

deepesn's People

Contributors

Stargazers

Watchers

deepesn's Issues

Application to simple multivariate time series forecast

Can I just use the multivariate piano for other datasets such as Jena climate dataset

ValueError when running main_MultivariatePolyphonic.py

D:\ProgramFilesNoSpace\Miniconda3\envs\py36\python.exe C:/Users/MONKEY/Desktop/DeepESN-master/main_MultivariatePolyphonic.py
Traceback (most recent call last):
  File "C:/Users/MONKEY/Desktop/DeepESN-master/main_MultivariatePolyphonic.py", line 80, in <module>
    main()
  File "C:/Users/MONKEY/Desktop/DeepESN-master/main_MultivariatePolyphonic.py", line 61, in main
    states = deepESN.computeState(dataset.inputs, deepESN.IPconf.DeepIP)
  File "C:\Users\MONKEY\Desktop\DeepESN-master\DeepESN\DeepESN.py", line 206, in computeState
    states = self.computeDeepIntrinsicPlasticity(inputs)
  File "C:\Users\MONKEY\Desktop\DeepESN-master\DeepESN\DeepESN.py", line 174, in computeDeepIntrinsicPlasticity
    self.computeLayerState(inputs[i], layer, 1)   
  File "C:\Users\MONKEY\Desktop\DeepESN-master\DeepESN\DeepESN.py", line 130, in computeLayerState
    state[:,0:1] = (1-self.lis[layer]) * initialStatesLayer + self.lis[layer] * np.tanh( np.multiply(self.Gain[layer], self.W[layer].dot(initialStatesLayer) + input[:,0:1]) + self.Bias[layer])        
ValueError: could not broadcast input array from shape (100,100) into shape (100,1)

Process finished with exit code 1

Accepting PR or updates ?

I recently found your repo after reading about DPS and models, and was curious if this repo is still active or if it's been deprecated

DeepESN for multivariate time-series classification problem

hello,

I would like to use this implementation for a multivariable time series classification.

My dataset is composed of several examples of 9 time-series segments with 128 steps.
Each segment belongs to one of 5 classes.

Input shape (2128,128,9) -> (number of examples, time-steps, variables)
output shape (2128,1) - > (number of examples, class)
(the class is one-hot-encoded so I use categorical-cross-entropy with 5 neurons output layer)

I have it working with a LSTMs based model.

I would like to test with deepESN to see if there is any improvement.

If you can help to provide some lights/base code to do this adaptation it would be nice.

Thanks in advance,
Rui

Self-Normalizing Activations on the Hyper-Sphere

Feature Request

consider including as an option a provision to apply a recently published new "spherical activation" function for ESNs, which is described in the paper below.

https://arxiv.org/abs/1903.11691
"Echo State Networks with Self-Normalizing Activations on the Hyper-Sphere"

lucapedrelli / deepesn Goto Github PK

deepesn's Introduction

Deep Echo State Network Python library (DeepESNpy), 2018

deepesn's People

Contributors

Stargazers

Watchers

Forkers

deepesn's Issues

Application to simple multivariate time series forecast

ValueError when running main_MultivariatePolyphonic.py

Accepting PR or updates ?

DeepESN for multivariate time-series classification problem

Self-Normalizing Activations on the Hyper-Sphere

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent