neos: End-to-End-Optimised Summary Statistics for High Energy Physics
neos: End-to-End-Optimised Summary Statistics for High Energy Physics
复制标题
neos：高能物理的端到端优化汇总统计
DOI：
--
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Lukas Heinrich
中科院分区：
文献类型：
--
作者：
Nathan Simpson;Lukas Heinrich
深度学习的出现产生了强大的工具来自动计算计算的梯度。这是因为训练神经网络等同于使用梯度下降迭代更新其参数以找到损失函数的最小值。深度学习是一个更广泛的范式的子集;一个具有自由参数的工作流程，是端到端可优化的，前提是人们可以一直跟踪梯度。本作品介绍了近地天体：遵循完全可微的高能物理工作流程的范例的示例实现，能够相对于分析的预期灵敏度优化可学习的汇总统计。这样做的结果是在一个优化过程中，是知道建模和处理系统的不确定性。
The advent of deep learning has yielded powerful tools to automatically compute gradients of computations. This is because training a neural network equates to iteratively updating its parameters using gradient descent to find the minimum of a loss function. Deep learning is then a subset of a broader paradigm; a workflow with free parameters that is end-to-end optimisable, provided one can keep track of the gradients all the way through. This work introduces neos: an example implementation following this paradigm of a fully differentiable high-energy physics workflow, capable of optimising a learnable summary statistic with respect to the expected sensitivity of an analysis. Doing this results in an optimisation process that is aware of the modelling and treatment of systematic uncertainties.