The cerebellum as a liquid state machine

Yamazaki, Tadashi; Tanaka, Shigeru

doi:10.1016/j.neunet.2007.04.004

Cited by 167 publications

(146 citation statements)

References 15 publications

Supporting

Mentioning

142

Contrasting

Order By: Relevance

“…The GR activity was a sparse representation of the input signal, so each simulation time sample corresponded to a different state of the granular layer. 31 The IOs received the US as a low-frequency random spike pattern 32 , not depending on the dynamics of the network but associated to the US event. The IOs were connected one by one to PCs through the Climbing Fibers (CFs).…”

Section: Computational Cerebellar Model and Optimizationmentioning

confidence: 99%

A Multiple-Plasticity Spiking Neural Network Embedded in a Closed-Loop Control System to Model Cerebellar Pathologies

Geminiani

Casellato

Antonietti

et al. 2018

Int. J. Neur. Syst.

View full text Add to dashboard Cite

The cerebellum plays a crucial role in sensorimotor control and cerebellar disorders compromise adaptation and learning of motor responses. However, the link between alterations at network level and cerebellar dysfunction is still unclear. In principle, this understanding would benefit of the development of an artificial system embedding the salient neuronal and plastic properties of the cerebellum and operating in closed-loop. To this aim, we have exploited a realistic spiking computational model of the cerebellum to analyse the network correlates of cerebellar impairment. The model was modified to reproduce three different damages of the cerebellar cortex: (i) a loss of the main output neurons (Purkinje Cells), (ii) a lesion to the main cerebellar afferents (Mossy Fibers), and (iii) a damage to a major mechanism of synaptic plasticity (Long Term Depression). The modified network models were challenged with an Eye-Blink Classical Conditioning test, a standard learning paradigm used to evaluate cerebellar impairment, in which the outcome was compared to reference results obtained in human or animal experiments. In all cases, the model reproduced the partial and delayed conditioning typical of the pathologies, indicating that an intact cerebellar cortex functionality is required to accelerate learning by transferring acquired information to the cerebellar nuclei. Interestingly, depending on the type of lesion, the redistribution of synaptic plasticity and response timing varied greatly generating specific adaptation patterns. Thus, not only the present work extends the generalization capabilities of the cerebellar spiking model to pathological cases, but also predicts how changes at the neuronal level are distributed across the network, making it usable to infer cerebellar circuit alterations occurring in cerebellar pathologies.

show abstract

Section: Computational Cerebellar Model and Optimizationmentioning

confidence: 99%

A Multiple-Plasticity Spiking Neural Network Embedded in a Closed-Loop Control System to Model Cerebellar Pathologies

Geminiani

Casellato

Antonietti

et al. 2018

Int. J. Neur. Syst.

View full text Add to dashboard Cite

show abstract

“…Functionally, a classification machine able to categorize any spatial pattern would show great divergence (in recent literature represented sparsity and decorrelations (Franzius et al 2007;Wyss et al 2003)), while a device transforming spatio in temporal patterns would show great convergence [a connectivity pattern that is at the core of echo-state networks and liquid-state machines (Yamazaki and Tanaka 2007)]. Should we be amazed that mechanistic hypotheses can be deduced from macroscopic features of the brain?…”

Section: Connection Schemes and Myelo-architectonicsmentioning

confidence: 99%

Valentino Braitenberg: From neuroanatomy to behavior and back

Negrello

2013

Biol Cybern

View full text Add to dashboard Cite

This article compiles an expose of Valentino Braitenberg's singular view on neuroanatomy and neuroscience. The review emphasizes his topologically informed work on neuroanatomy and his dialectics of brain-based explanations of motor behavior. Some of his early ideas on topologically informed neuroanatomy are presented, together with some of his more obscure work on the taxonomy of neural fiber bundles and synaptic arborizations. His functionally informed interpretations of neuroanatomy of the cerebellum, cortex, and hippocampus, are introduced. Finally, we will touch on his philosophical views and the inextricable role of function in the explanation of neural behavior.Keywords Neuroscience · Neuroanatomy · History · Braitenberg · Behavioral function ProlegomenaAbove great achievements, what brings the community to admire a scientist is a set of characteristics that transcend science and knowledge and approach nobility and wisdom. Science is a human endeavor with the commendable goal of understanding reality, in the hope that advancing science will bring us to better understand ourselves, and ourselves in the world. Through the work of great men, we approach that ideal.In great minds, tenacity is augmented by the ability to channel the product of observations with clarity and flair. In their descriptions the reality beneath the experiment becomes alive, unencumbered by jargon, wants to speak itself. The constraints imposed by experiment rather than nuisances, appear as scaffolds. For them, nature both suggests and takes precedence over theoretical constructs. As a consequence, their prose often reads as if it would have written itself, as if truth yearned for light.This article was conceived before the passing of Valentino Braitenberg, in the hope that he would have the opportunity to personally review it. Due to that it will likely at times read as an eulogy, but in that case, I believe, excusably. Braitenberg was a model of scientific attitude, both sober and jovial, inspired and humble, strict, and flexible. Simultaneously able to capture the essentials in an effortless glance, while swiftly accepting the burden of proof. With clarity, he indicated the flimsy, the flaky, the incomplete, the obscure, the messy, the missing, meanwhile stressing the necessary, the inescapable, the solid. Sternly, he offered pondered and polished logical foundations of his beliefs.The article attempts to characterize Valentino's manner of doing neuroanatomy and neuroscience while enlisting some of his most significant contributions to theory, method, and philosophy of the functional interpretation of neuroanatomy.With the hope of becoming great, we attempt to mirror ourselves against models. In neuroscience we have great men to mirror ourselves against. Cajal, Sherrington, Varela, Bernstein, von Helmholz, Reichardt, and now Braitenberg. Valentino Braitenberg was a living influence on many, and his legacy will be there for many to discover and learn from. 123

show abstract

“…Therefore, the states of this dynamic reservoir are linearly combined in an output layer, which is the sole trained part of the architecture. This type of state-dependent computation has been proposed as a biologically plausible model for cortical processing (Buonomano and Maass, 2009;Maass et al, 2002;Yamazaki and Tanaka, 2007). Such theoretical models include: Echo State Networks (Jaeger and Haas, 2004) for analog neurons and Liquid State Machines (Maass et al, 2002) for spiking neurons.…”

Section: Introductionmentioning

confidence: 99%

Learning slow features with reservoir computing for biologically-inspired robot localization

Antonelo

Schrauwen

2012

Neural Networks

View full text Add to dashboard Cite

This work proposes a hierarchical biologically-inspired architecture for learning sensor-based spatial representations of a robot environment in an unsupervised way. The first layer is comprised of a fixed randomly generated recurrent neural network, the reservoir, which projects the input into a high-dimensional, dynamic space. The second layer learns instantaneous slowly-varying signals from the reservoir states using Slow Feature Analysis (SFA), whereas the third layer learns a sparse coding on the SFA layer using Independent Component Analysis (ICA). While the SFA layer generates non-localized activations in space, the ICA layer presents high place selectivity, forming a localized spatial activation, characteristic of place cells found in the hippocampus area of the rodent's brain. We show that, using a limited number of noisy short-range distance sensors as input, the proposed system learns a spatial representation of the environment which can be used to predict the actual location of simulated and real robots, without the use of odometry. The results confirm that the reservoir layer is essential for learning spatial representations from low-dimensional input such as distance sensors. The main reason is that the reservoir state reflects the recent history of the input stream. Thus, this fading memory is essential for detecting locations, mainly when locations are ambiguous and characterized by similar sensor readings.

show abstract

The cerebellum as a liquid state machine

Cited by 167 publications

References 15 publications

A Multiple-Plasticity Spiking Neural Network Embedded in a Closed-Loop Control System to Model Cerebellar Pathologies

A Multiple-Plasticity Spiking Neural Network Embedded in a Closed-Loop Control System to Model Cerebellar Pathologies

Valentino Braitenberg: From neuroanatomy to behavior and back

Learning slow features with reservoir computing for biologically-inspired robot localization

Contact Info

Product

Resources

About