美国华盛顿大学Sam A. Golden和Simon R. O. Nilsson共同合作，近期取得重要工作进展。他们研究提出将简单行为分析（SimBA）作为行为神经科学中可解释机器学习的平台。相关研究成果2024年5月22日在线发表于《自然—神经科学》杂志上。

据介绍，由于缺乏可量化的行为定义和行为注释的主观性质，在使用手动注释时，对复杂行为的研究往往具有挑战性。监督机器学习方法的集成通过包含可访问和可解释的模型解释来减轻其中一些问题。

为了减少访问障碍，并加强可访问模型的可解释性，研究人员为行为神经科学家开发了开源的简单行为分析（SimBA）平台。SimBA引入了几种机器学习可解释性工具，包括SHAPLEY Additive exPlanation (SHAP)评分，这些工具有助于创建可解释和透明的行为分类器。研究人员展示了可解释性指标的添加如何允许，对研究群体和物种之间的攻击性社会行为进行量化比较，将行为重新定义为可共享的资源，并提供一个开源框架。

总之，这一研究提供了一个开源的、图形用户界面（GUI）驱动的、文档齐全的包，以促进实验室之间行为分类工具的自动化和共享。

附：英文原文

Title: Simple Behavioral Analysis (SimBA) as a platform for explainable machine learning in behavioral neuroscience

Author: Goodwin, Nastacia L., Choong, Jia J., Hwang, Sophia, Pitts, Kayla, Bloom, Liana, Islam, Aasiya, Zhang, Yizhe Y., Szelenyi, Eric R., Tong, Xiaoyu, Newman, Emily L., Miczek, Klaus, Wright, Hayden R., McLaughlin, Ryan J., Norville, Zane C., Eshel, Neir, Heshmati, Mitra, Nilsson, Simon R. O., Golden, Sam A.

Issue&Volume: 2024-05-22

Abstract: The study of complex behaviors is often challenging when using manual annotation due to the absence of quantifiable behavioral definitions and the subjective nature of behavioral annotation. Integration of supervised machine learning approaches mitigates some of these issues through the inclusion of accessible and explainable model interpretation. To decrease barriers to access, and with an emphasis on accessible model explainability, we developed the open-source Simple Behavioral Analysis (SimBA) platform for behavioral neuroscientists. SimBA introduces several machine learning interpretability tools, including SHapley Additive exPlanation (SHAP) scores, that aid in creating explainable and transparent behavioral classifiers. Here we show how the addition of explainability metrics allows for quantifiable comparisons of aggressive social behavior across research groups and species, reconceptualizing behavior as a sharable reagent and providing an open-source framework. We provide an open-source, graphical user interface (GUI)-driven, well-documented package to facilitate the movement toward improved automation and sharing of behavioral classification tools across laboratories.

DOI: 10.1038/s41593-024-01649-9

Source: https://www.nature.com/articles/s41593-024-01649-9