OSBAD
收藏资源简介:
OSBAD是一个开放源代码的异常检测基准,用于电池应用中的异常检测框架。它包含15种不同的算法,包括统计、基于距离和无监督机器学习方法。通过在具有不同电极材料和电解质组成的两个独立数据集上评估OSBAD,展示了其跨化学泛化能力,以识别不同电化学系统中的不规则性。通过提供具有开源、可重复的异常检测工作流程的基准测试数据库,OSBAD为开发安全的、可扩展的和可迁移的异常检测工具建立了统一的框架。
OSBAD is an open-source anomaly detection benchmark designed for anomaly detection frameworks in battery applications. It includes 15 distinct algorithms, covering statistical, distance-based, and unsupervised machine learning methods. Its cross-chemical generalization capability for identifying irregularities across diverse electrochemical systems was demonstrated by evaluating OSBAD on two independent datasets with varying electrode materials and electrolyte compositions. By providing a benchmark database paired with an open-source and reproducible anomaly detection workflow, OSBAD establishes a unified framework for developing safe, scalable, and transferable anomaly detection tools.

- 1通过BASF SE, Ludwigshafen am Rhein, Germany; Institute of Multidisciplinary Research for Advanced Materials, Tohoku University · 2025年



