Developing a structural-based local learning rule for classification tasks using ionic liquid space-based reservoir

Iranmehr, Ensieh; Shouraki, Saeed Bagheri; Faraji, Mohammad Mahdi

doi:10.1007/s00521-022-07345-8

Cited by 3 publications

(3 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, investigating whether the findings of this study can be applied not only to simple single-layer reservoirs but also to reservoirs of various structures, such as multi-reservoirs 15 or self-modulated RC 16 , is also a challenge. Additionally, when applying our findings to models with spiking neurons models in LSM 18 , 19 , it is crucial to highlight the challenges, because the delay capacity used in our study may not be directly applicable in the case with spiking neurons. Therefore, developing alternative metrics for assessing similar characteristics is imperative.…”

Section: Discussionmentioning

confidence: 99%

“…Additionally, Iinuma et al found that parallelisation of reservoir assembly is effective for tasks requiring multi-dimensional inputs 14 . Reservoirs with various neuronal dynamics characteristics, including ESN and liquid state machines (LSMs), have been extensively studied, as shown in the literatures 18 , 19 . In terms of intrinsic neuronal dynamics, research has highlighted the importance of fine-tuning the neurons’ time-history terms to optimise the time scale of neuronal activity.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Impact of time-history terms on reservoir dynamics and prediction accuracy in echo state networks

Ebato,

Nobukawa,

Sakemi

et al. 2024

Sci Rep

View full text Add to dashboard Cite

The echo state network (ESN) is an excellent machine learning model for processing time-series data. This model, utilising the response of a recurrent neural network, called a reservoir, to input signals, achieves high training efficiency. Introducing time-history terms into the neuron model of the reservoir is known to improve the time-series prediction performance of ESN, yet the reasons for this improvement have not been quantitatively explained in terms of reservoir dynamics characteristics. Therefore, we hypothesised that the performance enhancement brought about by time-history terms could be explained by delay capacity, a recently proposed metric for assessing the memory performance of reservoirs. To test this hypothesis, we conducted comparative experiments using ESN models with time-history terms, namely leaky integrator ESNs (LI-ESN) and chaotic echo state networks (ChESN). The results suggest that compared with ESNs without time-history terms, the reservoir dynamics of LI-ESN and ChESN can maintain diversity and stability while possessing higher delay capacity, leading to their superior performance. Explaining ESN performance through dynamical metrics are crucial for evaluating the numerous ESN architectures recently proposed from a general perspective and for the development of more sophisticated architectures, and this study contributes to such efforts.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Impact of time-history terms on reservoir dynamics and prediction accuracy in echo state networks

Ebato,

Nobukawa,

Sakemi

et al. 2024

Sci Rep

View full text Add to dashboard Cite

show abstract

“…On the other hand, approaches based on synaptic plasticity [31][32][33][34] use supervised or reinforcement learning principles to adjust the parameters of the model in real-time. Such online learning methods rely on retraining the model, which is impractical in real-world applications.…”

Section: Literature Reviewmentioning

confidence: 99%

Synaptic plasticity-based regularizer for artificial neural networks

Yousef,

2024

Preprint

View full text Add to dashboard Cite

Regularization is an important tool for the generalization of ANN models. Due to the lack of constraints, it cannot guarantee that the model will work in a real environment with continuous changes in the distribution. Inspired by neuroplasticity, this paper proposes a bounded regularization method that can be safely activated during the deployment phase. First, we improve the reliability of the outputs of selected neurons by extending our recently proposed neuronal masking. Subsequently, we regularize the model by introducing a synaptic connection module to determine the connection of the masks to their previous layer based on the coming input data. To find the optimal connection, we define a mixed-integer nonlinear programming (MINLP) problem to minimize the loss of prospect uncertainty and solve it using our proposed “single wave” method. Finally, we propose a storage/recovery memory module to memorize these connections along with the corresponding uncertainty level. Experimental results from classification and regression tasks show that the proposed method outperforms the state-of-the-art in the sense of accuracy.

show abstract

Unsupervised Extraction of Shape-Based Signal Patterns for Incoming Signal Recognition

et al. 2023

View full text Add to dashboard Cite

Developing a structural-based local learning rule for classification tasks using ionic liquid space-based reservoir

Cited by 3 publications

References 62 publications

Impact of time-history terms on reservoir dynamics and prediction accuracy in echo state networks

Impact of time-history terms on reservoir dynamics and prediction accuracy in echo state networks

Synaptic plasticity-based regularizer for artificial neural networks

Unsupervised Extraction of Shape-Based Signal Patterns for Incoming Signal Recognition

Contact Info

Product

Resources

About