The lack of evidence for new physics at the Large Hadron Collider so far has prompted the development of model-independent search techniques. In this study, we compare the anomaly scores of a variety of anomaly detection techniques: an isolation forest, a Gaussian mixture model, a static autoencoder, and a β-variational autoencoder (VAE), where we define the reconstruction loss of the latter as a weighted combination of regression and classification terms. We apply these algorithms to the 4-vectors of simulated LHC data, but also investigate the performance when the non-VAE algorithms are applied to the latent space variables created by the VAE. In addition, we assess the performance when the anomaly scores of these algorithms are combined in various ways. Using super- symmetric benchmark points, we find that the logical AND combination of the anomaly scores yielded from algorithms trained in the latent space of the VAE is the most effective discriminator of all methods tested.

Charged Higgs bosons are predicted in variety of theoretically well-motivated new physics models with extended Higgs sectors. In this study, we focus on a type-II two Higgs doublet model (2HDM-II) and consider a heavy charged Higgs with its mass ranging from 500 GeV to 1 TeV as dictated by the b → sγ constraints which render M H ± > 480 GeV. We study the dominant production mode H ± t associated production with H ± → W ± A being the dominant decay channel when the pseudoscalar A is considerably lighter. For such a heavy charged Higgs, both the decay products W ± and A are relatively boosted. In such a scenario, we apply the jet substructure analysis of tagging the fat pseudoscalar and W jets in order to eliminate the standard model background efficiently. We perform a detailed detector simulation for the signal and background processes at the 14 TeV LHC. We introduce various kinematical cuts to determine the signal significance for a number of benchmark points with charged Higgs boson mass from 500 GeV to 1 TeV in the W ± A decay channel. Finally we perform a multivariate analysis utilizing a boosted decision tree algorithm to optimize these significances.

In the framework of the type-II Two Higgs Doublet Model (2HDM-II) a charged Higgs search strategy is presented for the dominant production mode gb → tH ± at the 14 TeV LHC. We consider the decay process which includes t → bW ± and H ± → AW ± , and a fully hadronic final state consisting of bbb + jets + X. Dictated by the b → sγ constraints which render M H ± > 480 GeV we study two scenarios in which the charged Higgs mass is 750 GeV and the pseudoscalar Higgs mass is 200 GeV and 500 GeV. In this mass scheme highly boosted final state objects are expected and handled with jet substructure techniques which also acts to suppress the standard model background. A detailed detector analysis is performed, followed by a multivariate analysis involving many kinematic variables to optimize signal to background significance. Finally the LHC search sensitivities for the two scenarios are presented for various integrated luminosities.