Training in Methods of Computational Neuroscience (MCN)

摘要 伍兹海洋生物实验室的计算神经科学（MCN）课程方法 孔，MA提供了一个独特的培训机会，使学生掌握尖端技能和概念 为理解大脑的生物学功能和神经机制奠定了计算方法的基础 细胞、电路和行为。MCN是一个密集的讲座和计算机实验室课程，针对有才华的 学生在生物和数学科学方面具有互补的专业知识。MCN学员互动 与国际知名科学家一对一，并与广泛的 各种计算技术和方法。MCN课程的目标是培养学生1） 在理解大脑功能的计算方法的概念和方法中， 神经数据分析的数学基础和技术3）使用计算工具 解决神经生物学中的基本问题，4）使用计算方法来诊断和 治疗人类大脑疾病，5）促进跨学科合作， 互动和指导。该课程提供了关于如何使用计算 方法来解决神经生物学的基本问题，使用实验技术应用于整个 从分子到行为的广泛生物尺度。这是通过以下方面的实践培训来实现的： 神经网络的数学基础和基于模型的统计和理论分析技术 神经元、网络和大脑系统的机械描述背景下的数据。MCN课程 不断更新与新的实验进展相关的计算主题，如解剖学 显微结构和全脑范围内的连通性，使用电子和光学的高密度测量 记录和使用光遗传学的神经回路的因果扰动。这些新的主题包括状态空间 轨迹分析，深度网络，压缩感知，高级统计推断，流形学习， 降维，分析大型数据集，包括全脑成像和大规模细胞类型 具体的连通性。另一个创新是越来越强调计算方法， 理解和解决大脑疾病，包括计算精神病学的新兴领域。 该课程的核心是与一群敬业的教师进行激烈互动的环境，其中许多人 他们花了几周甚至一个月的时间和学生们一起工作。在课程的后半部分， 专注于由教师和教学团队指导的新颖尖端研究项目 助手该课程通过以下方式促进学生职业生涯的国际化环境 为学生提供网络机会，导致长期的互动，指导， 合作。

ABSTRACT The Methods in Computational Neuroscience (MCN) course at the Marine Biological Laboratory in Woods Hole, MA provides a unique training opportunity to equip students with cutting edge skills and conceptual foundations of computational methods to understanding the biological function and neural mechanisms of brain cells, circuits, and behavior. MCN is an intense lecture and computer laboratory course targeted to a talented mix of students with complementary expertise in biological and mathematical sciences. MCN students interact one-on-one with internationally renowned scientists and carry out hands-on, project-based training with a wide variety of computational techniques and approaches. The goals of the MCN course are to train students 1) in the concepts and methods of computational approaches to understanding brain function 2) in the mathematical foundations and techniques for the analysis of neural data 3) to use computational tools to solve fundamental questions in neurobiology, 4) to use computational methods to diagnose and treat human brain disorders and 5) to foster interdisciplinary collaborations that lead to long term interactions and mentorship. The course provides in-depth instruction on how to use computational approaches to solve fundamental problems in neurobiology using experimental techniques applied across a wide range of biological scales from molecules to behavior. This is accomplished through hands-on training in the mathematical foundations and techniques for the model-based, statistical and theoretical analysis of neural data in the context of mechanistic descriptions of neurons, networks, and brain systems. The MCN course continually updates its computational topics that are relevant to new experimental advances, such anatomical connectivity at microstructural and brain-wide scales, high density measurements using electrical and optical recording, and causal perturbations of neural circuits using optogenetics. These new topics include state-space trajectory analysis, deep networks, compressed sensing, advanced statistical inference, manifold learning, dimensionality reduction, analysis of large data sets including brain wide imaging and large scale cell-type specific connectivity. An additional innovation is an increased emphasis on computational approaches to understanding and addressing brain disorders, including the emerging field of computational psychiatry. Central to the course is an environment of intense interactions with a group of dedicated faculty, many of whom devote weeks or even the entire month to work with students. In the last half of the course students devote concentrated effort to novel cutting edge research projects mentored by a team of faculty and teaching assistants. The course fosters a collegial, international environment that advances student careers by providing students with networking opportunities that lead to long-term interactions, mentorship, and collaborations.