Learning probabilistic neural representations with randomly connected circuits.
Learning probabilistic neural representations with randomly connected circuits.
复制标题
DOI：
10.1073/pnas.1912804117
复制
发表时间：
2020-10-06
影响因子：
11.1
通讯作者：
Schneidman E
中科院分区：
文献类型：
--
作者：
Maoz O;Tkačik G;Esteki MS;Kiani R;Schneidman E
受真实的神经回路的随机连接性和随机投影的数学能力的启发，我们提出了神经回路的设计和功能的理论。具体来说，我们介绍了一个家庭的大型神经群体代码的统计模型，一个简单的神经电路架构，将实现这些模型，和一个生物学上合理的学习规则，这样的电路。由此产生的神经结构表明了神经回路的设计原则，即它们在没有明确的教学信号的情况下，在给定过去的输入的情况下，学习计算其输入的数学惊喜。我们将这些模型应用于猴子视觉和前额叶皮层中大型神经群体的记录，并显示它们是高度准确，高效和可扩展的。大脑使用一种嘈杂的、基于尖峰的神经代码来表示和概率性地推理复杂的刺激和运动动作。这种神经计算的一个关键构建块，以及监督和无监督学习的基础，是估计输入的高维神经活动模式的惊喜或可能性的能力。尽管在神经反应和深度学习的统计建模方面取得了进展，但目前的方法要么无法扩展到大型神经群体，要么无法使用生物现实机制来实现。受真实的神经元回路的稀疏和随机连接的启发，我们提出了一个神经代码模型，该模型准确估计了个体尖峰模式的可能性，并且具有简单、可扩展、高效、可学习和现实的神经实现。该模型在同时记录猴视觉和前额叶皮层中>100个神经元的尖峰活动方面的性能与最先进的模型相当或更好。重要的是，可以使用少量样本并使用利用神经电路固有噪声的局部学习规则来学习模型。随机连接的结构变化更慢，与重新布线和修剪过程一致，进一步提高了所产生的神经表示的效率和稀疏性。我们的研究结果融合了神经解剖学，机器学习和理论神经科学的见解，建议随机稀疏连接作为神经元计算的关键设计原则。
We present a theory of neural circuits’ design and function, inspired by the random connectivity of real neural circuits and the mathematical power of random projections. Specifically, we introduce a family of statistical models for large neural population codes, a straightforward neural circuit architecture that would implement these models, and a biologically plausible learning rule for such circuits. The resulting neural architecture suggests a design principle for neural circuit—namely, that they learn to compute the mathematical surprise of their inputs, given past inputs, without an explicit teaching signal. We applied these models to recordings from large neural populations in monkeys’ visual and prefrontal cortices and show them to be highly accurate, efficient, and scalable. The brain represents and reasons probabilistically about complex stimuli and motor actions using a noisy, spike-based neural code. A key building block for such neural computations, as well as the basis for supervised and unsupervised learning, is the ability to estimate the surprise or likelihood of incoming high-dimensional neural activity patterns. Despite progress in statistical modeling of neural responses and deep learning, current approaches either do not scale to large neural populations or cannot be implemented using biologically realistic mechanisms. Inspired by the sparse and random connectivity of real neuronal circuits, we present a model for neural codes that accurately estimates the likelihood of individual spiking patterns and has a straightforward, scalable, efficient, learnable, and realistic neural implementation. This model’s performance on simultaneously recorded spiking activity of >100 neurons in the monkey visual and prefrontal cortices is comparable with or better than that of state-of-the-art models. Importantly, the model can be learned using a small number of samples and using a local learning rule that utilizes noise intrinsic to neural circuits. Slower, structural changes in random connectivity, consistent with rewiring and pruning processes, further improve the efficiency and sparseness of the resulting neural representations. Our results merge insights from neuroanatomy, machine learning, and theoretical neuroscience to suggest random sparse connectivity as a key design principle for neuronal computation.
登录
查看更多内容
影响因子：
25
作者：
Goris, Robbe L. T.;Movshon, J. Anthony;Simoncelli, Eero P.
通讯作者：
Simoncelli, Eero P.
DOI：
10.1073/pnas.1019641108
发表时间：
2011-06-07
影响因子：
11.1
作者：
Ganmor, Elad;Segev, Ronen;Schneidman, Elad
通讯作者：
Schneidman, Elad
DOI：
10.1523/jneurosci.2753-12.2013
发表时间：
2013-02-27
期刊：
The Journal of neuroscience : the official journal of the Society for Neuroscience
影响因子：
--
作者：
Barak O;Rigotti M;Fusi S
通讯作者：
Fusi S
影响因子：
2.5
作者：
Candes, Emmanuel J.;Eldar, Yonina C.;Randall, Paige
通讯作者：
Randall, Paige
影响因子：
5.3
作者：
El-Shamayleh, Yasmine;Kumbhani, Romesh D.;Movshon, J. Anthony
通讯作者：
Movshon, J. Anthony