摘要
由一名新闻记者兼机器人与机器学习每日新闻编辑-调查人员讨论机器学习的新发现。根据News Rx编辑在德国卡尔斯鲁厄的新闻报道，研究表明：“如果我们最喜欢的计算工具——无论是数字、统计、机器学习方法，还是任何计算机模型只提供单值输出，我们如何量化不确定性？在这篇文章中，我们引入了简单的不确定性量化(EasyUQ)技术，它将实值模型输出转化为校准的统计分布。”EasyUQ完全基于模型输出-结果对的训练数据，不需要输入SS模型。EasyUQ的基本形式是最近推出的等渗分布回归(IDR)技术的一个特例，该技术利用POO L-邻接违反者算法进行非参数等渗回归。这项研究的财政支持者包括德国研究基金会(DFG)、瑞士国家科学基金会(SNSF)、克劳斯·齐拉基金会。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – Investigators discuss new findings in Machine Learning. According to news reporting out of Karlsruhe, Germany, by News Rx editors, research stated, “How can we quantify uncertainty if our favorite co mputational tool—be it a numerical, statistical, or machine learning approach, or just any computer model-provides singlevalued output only? In this article, we introduce the Easy Uncertainty Quantification (EasyUQ) technique, which trans forms real-valued model output into calibrated statistical distributions, based solely on training data of model output–outcome pairs, without any need to acce ss model input. In its basic form, EasyUQ is a special case of the recently intr oduced isotonic distributional regression (IDR) technique that leverages the poo l-adjacent-violators algorithm for nonparametric isotonic regression.” Financial supporters for this research include German Research Foundation (DFG), Swiss National Science Foundation (SNSF), Klaus Tschira Foundation.
NETL
NSTL