higgs

  • Description:

The data has been produced using Monte Carlo simulations. The first 21 features (columns 2-22) are kinematic properties measured by the particle detectors in the accelerator. The last seven features are functions of the first 21 features; these are high-level features derived by physicists to help discriminate between the two classes. There is an interest in using deep learning methods to obviate the need for physicists to manually develop such features. Benchmark results using Bayesian Decision Trees from a standard physics package and 5-layer neural networks are presented in the original paper.

Split Examples
'train' 11,000,000
  • Feature structure:
FeaturesDict({
    'class_label': float32,
    'jet_1_b-tag': float64,
    'jet_1_eta': float64,
    'jet_1_phi': float64,
    'jet_1_pt': float64,
    'jet_2_b-tag': float64,
    'jet_2_eta': float64,
    'jet_2_phi': float64,
    'jet_2_pt': float64,
    'jet_3_b-tag': float64,
    'jet_3_eta': float64,
    'jet_3_phi': float64,
    'jet_3_pt': float64,
    'jet_4_b-tag': float64,
    'jet_4_eta': float64,
    'jet_4_phi': float64,
    'jet_4_pt': float64,
    'lepton_eta': float64,
    'lepton_pT': float64,
    'lepton_phi': float64,
    'm_bb': float64,
    'm_jj': float64,
    'm_jjj': float64,
    'm_jlv': float64,
    'm_lv': float64,
    'm_wbb': float64,
    'm_wwbb': float64,
    'missing_energy_magnitude': float64,
    'missing_energy_phi': float64,
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
class_label Tensor float32
jet_1_b-tag Tensor float64
jet_1_eta Tensor float64
jet_1_phi Tensor float64
jet_1_pt Tensor float64
jet_2_b-tag Tensor float64
jet_2_eta Tensor float64
jet_2_phi Tensor float64
jet_2_pt Tensor float64
jet_3_b-tag Tensor float64
jet_3_eta Tensor float64
jet_3_phi Tensor float64
jet_3_pt Tensor float64
jet_4_b-tag Tensor float64
jet_4_eta Tensor float64
jet_4_phi Tensor float64
jet_4_pt Tensor float64
lepton_eta Tensor float64
lepton_pT Tensor float64
lepton_phi Tensor float64
m_bb Tensor float64
m_jj Tensor float64
m_jjj Tensor float64
m_jlv Tensor float64
m_lv Tensor float64
m_wbb Tensor float64
m_wwbb Tensor float64
missing_energy_magnitude Tensor float64
missing_energy_phi Tensor float64
  • Citation:
@article{Baldi:2014kfa,
      author         = "Baldi, Pierre and Sadowski, Peter and Whiteson, Daniel",
      title          = "{Searching for Exotic Particles in High-Energy Physics
                        with Deep Learning}",
      journal        = "Nature Commun.",
      volume         = "5",
      year           = "2014",
      pages          = "4308",
      doi            = "10.1038/ncomms5308",
      eprint         = "1402.4735",
      archivePrefix  = "arXiv",
      primaryClass   = "hep-ph",
      SLACcitation   = "%%CITATION = ARXIV:1402.4735;%%"
}