arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2108.07537cs.LG
收藏

用柔性样条基估计光滑稀疏的神经感受野

Estimating smooth and sparse neural receptive fields with a flexible spline basis

Institute for Ophthalmic Research, University of Tübingen(图宾根大学眼科研究所);Centre for Integrative Neuroscience, University of Tübingen(图宾根大学综合神经科学中心);Tübingen AI Center, University of Tübingen(图宾根大学图宾根人工智能中心)

Ziwei Huang, Yanli Ran, Jonathan Oesterle, Thomas Euler, Philipp Berens

摘要:

时空感受野(STRF)模型经常被用来近似由感觉神经元执行的计算。通常,这样的STRF被假定为平滑和稀疏的。目前的国家的最先进的方法来估计基于经验贝叶斯的STRF通常是不计算效率在高维设置中,如在感觉神经科学中遇到的。在这里,我们追求一种替代方法,并通过选择一组具有所需属性的基函数(自然三次样条)来编码用于估计STRF的先验知识。我们的方法是计算效率高,可以很容易地应用到广泛的现有模型。我们比较了基于样条的方法,非样条的模拟和实验数据的性能,表明基于样条的方法始终优于非样条版本。

英文摘要:

Spatio-temporal receptive field (STRF) models are frequently used to approximate the computation implemented by a sensory neuron. Typically, such STRFs are assumed to be smooth and sparse. Current state-of-the-art approaches for estimating STRFs based on empirical Bayes are often not computationally efficient in high-dimensional settings, as encountered in sensory neuroscience. Here we pursued an alternative approach and encode prior knowledge for estimation of STRFs by choosing a set of basis functions with the desired properties: natural cubic splines. Our method is computationally efficient and can be easily applied to a wide range of existing models. We compared the performance of spline-based methods to non-spline ones on simulated and experimental data, showing that spline-based methods consistently outperform the non-spline versions.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑