"Figure 6 - right, high pt DeepTau vs ele WPs" of "Identification of hadronic tau lepton decays using a deep neural network"

Efficiency for electrons against efficiency for genuine tau_h to pass the MVA and De discriminators, separately for electrons and tau_h with 20 < pt < 100 GeV (left) and pt > 100 GeV (right). Vertical bars correspond to the statistical uncertainties. The tau_h candidates are reconstructed in one of the tau_h decay modes without missing charged hadrons. Compared with the MVA discriminator, the De discriminator reduces the electron efficiency by more than a factor of two for a tau_h efficiency of 70% and by more than a factor of 10 for τh efficiencies larger than 88%. Furthermore, working points (indicated as full circles) are now provided for previously inaccessible tau_h efficiencies larger than 90%, for a misidentification efficiency between 0.3 and 8%.

分别针对横动量（transverse momentum，pt）满足20 < pt < 100 GeV（左图）以及pt > 100 GeV（右图）的电子与τ_h（tau_h）粒子，对比展示电子通过MVA与De判别器的效率，以及真实τ_h通过两类判别器的效率。图中的垂直误差棒代表统计不确定性。本次分析中的τ_h候选粒子均通过不含丢失带电强子的τ_h衰变道重建得到。相较于MVA判别器，当τ_h识别效率为70%时，De判别器可将电子的误识别效率降至原水平的1/2以下；而当τ_h识别效率超过88%时，电子误识别效率的降低幅度更是超过一个量级。此外，本次研究还提供了此前难以实现的、τ_h识别效率超过90%且电子误识别效率处于0.3%至8%区间的工作点（以实心圆点标示）。