CMS Phase 1 heavy flavour identification performance and developments

At the Large Hadron Collider, the identification of jets originating from heavy flavour quarks (b or c tagging) is important for searches for new physics and for measurements of standard model processes. A variety of b tagging algorithms has been developed at CMS to select b-quark jets based on variables such as the impact parameters of the charged-particle tracks, the properties of reconstructed decay vertices, and the presence or absence of a lepton, or combinations thereof. The CMS Phase 1 upgrade includes a new pixel detector with an addition layer. In consequence, the resolution of the track reconstruction is expected to improve, in particular close to the beam spot, leading to a more precise determination of the variables important for b tagging. The expected effect on the b tagging performance is presented for a selected set of existing algorithms: CSVv2, cMVAv2, and DeepCSV. The latter deep neural network based multi-classification algorithm is re-trained based on the new detector geometry, while the others do not benefit from a retraining. In addition, the performance of a new multi-classification algorithm under development, the DeepFlavour algorithm, is presented. In contrast to the other algorithms, it uses properties of charged and neutral particle-flow candidates, as well as of secondary vertices within the jet, without a b tagging specific preselection. The neural network consists of multiple 1x1 convolutional layers for each input collection, followed by several densely connected layers. In simulation this algorithm outperforms the other taggers significantly, in particular for high-pt jets, which can, e.g., lead to improved sensitivity in searches for new physics with high energetic b jets in the final state.