Low energy event classification in IceCube using boosted decision trees

2021 ◽  
Vol 16 (12) ◽  
pp. C12007
Author(s):  
K. Leonard DeHolton
Keyword(s):  
Cherenkov Radiation ◽  
Atmospheric Neutrino ◽  
Particle Identification ◽  
Charged Current ◽  
Event Classification ◽  
Oscillation Analysis ◽  
Traditional Classification ◽  
Low Energies ◽  
Boosted Decision Tree ◽  
Boosted Decision Trees

Abstract The DeepCore sub-array within the IceCube Neutrino Observatory is a densely instrumented region of Antarctic ice designed to observe atmospheric neutrino interactions above 5 GeV via Cherenkov radiation. An essential aspect of any neutrino oscillation analysis is the ability to accurately identify the flavor of neutrino events in the detector. This task is particularly difficult at low energies when very little light is deposited in the detector. Here we discuss the use of machine learning to perform event classification at low energies in IceCube using a boosted decision tree (BDT). A BDT is trained using reconstructed quantities to identify track-like events, which result from muon neutrino charged current interactions. This new method improves the accuracy of particle identification compared to traditional classification methods which rely on univariate straight cuts.

RECENT RESULTS FROM K2K

2002 ◽  
Vol 17 (24) ◽  
pp. 3364-3377 ◽  
Author(s):  
◽  
C. K. JUNG
Keyword(s):  
Neutrino Oscillation ◽  
Atmospheric Neutrino ◽  
Neutrino Beam ◽  
Statistical Fluctuation ◽  
Atmospheric Neutrinos ◽  
Long Baseline ◽  
Neutrino Oscillation Experiment ◽  
Oscillation Analysis ◽  
The Us ◽  
Oscillation Experiment

K2K is a long baseline neutrino oscillation experiment using a neutrino beam produced at the KEK 12 GeV PS, a near detector complex at KEK and a far detector (Super-Kamiokande) in Kamioka, Japan. The experiment was constructed and is being operated by an international consortium of institutions from Japan, Korea, and the US. The experiment started taking data in 1999 and has successfully taken data for about two years. K2K is the first long beseline neutrino oscillation experiment with a baseline of order hundreds of km and is the first accelerator based neutrino oscillation experiment that is sensitive to the Super-Kamiokande allowed region obtained from the atmospheric neutrino oscillation analysis. A total of 44 events have been observed in the far detector during the period of June 1999 to April 2001 corresponding to 3.85 × 1019 protons on target. The observation is consistent with the neutrino oscillation expectations based on the oscillation parameters derived from the atmospheric neutrinos, and the probability that this is a statistical fluctuation of non-oscillation expectation of [Formula: see text] is less than 3%.

scholarly journals Particle identification using Boosted Decision Trees in the Semi-Digital Hadronic Calorimeter prototype

2020 ◽  
Vol 15 (10) ◽  
pp. P10009-P10009
Author(s):  
D. Boumediene ◽  
A. Pingault ◽  
M. Tytgat ◽  
B. Bilki ◽  
D. Northacker ◽  
...  
Keyword(s):  
Decision Trees ◽  
Particle Identification ◽  
Hadronic Calorimeter ◽  
Boosted Decision Trees

scholarly journals Comparison of atmospheric neutrino flux calculations at low energies

1996 ◽  
Vol 54 (9) ◽  
pp. 5578-5584 ◽  
Author(s):  
T. K. Gaisser ◽  
M. Honda ◽  
K. Kasahara ◽  
H. Lee ◽  
S. Midorikawa ◽  
...  
Keyword(s):  
Atmospheric Neutrino ◽  
Neutrino Flux ◽  
Low Energies

RESULTS FROM THE PURE D2O PHASE OF THE SUDBURY NEUTRINO OBSERVATORY

2003 ◽  
Vol 18 (22) ◽  
pp. 3789-3807
Author(s):  
◽  
F. A. DUNCAN
Keyword(s):  
Neutral Current ◽  
Solar Neutrinos ◽  
Total Flux ◽  
Sudbury Neutrino Observatory ◽  
Charged Current ◽  
Mixing Angle ◽  
Cerenkov Detector ◽  
Neutrino Spectrum ◽  
Oscillation Analysis ◽  
Current Interaction

The Sudbury Neutrino Observatory is a 1000 T D2O Cerenkov detector that is sensitive to 8 B and hep solar neutrinos. Both Charged Current and Neutral Current interaction rates on deuterons as well as the Elastic Scattering interaction rate on electrons can be measured simultaneously. Assuming an undistorted 8 B neutrino spectrum, the total flux measured with the NC reaction is [Formula: see text], which is consistent with solar models. The νe component of the 8 B solar flux is [Formula: see text] for a kinetic energy threshold of 5 MeV. The non-νe component is [Formula: see text], which is 5.3σ greater than zero, giving strong evidence for solar νe flavor transformation. The Day-Night Asymmetry for the Charged Current interaction is [Formula: see text]. If the total flux of active neutrinos is additionally constrained to have no asymmetry, the νe asymmetry is found to be [Formula: see text]. Combined with other solar neutrino data, a global MSW oscillation analysis strongly favors the Large Mixing Angle (LMA) solution.

Advances in particle identification in AMS at low energies

2007 ◽  
Vol 259 (1) ◽  
pp. 165-172 ◽  
Author(s):  
Martin Suter ◽  
Max Döbeli ◽  
Michal Grajcar ◽  
Arnold Müller ◽  
Martin Stocker ◽  
...  
Keyword(s):  
Particle Identification ◽  
Low Energies

scholarly journals Nanosecond machine learning event classification with boosted decision trees in FPGA for high energy physics

2021 ◽  
Vol 16 (08) ◽  
pp. P08016
Author(s):  
T.M. Hong ◽  
B.T. Carlson ◽  
B.R. Eubanks ◽  
S.T. Racz ◽  
S.T. Roche ◽  
...  
Keyword(s):  
Machine Learning ◽  
Decision Trees ◽  
High Energy Physics ◽  
High Energy ◽  
Event Classification ◽  
Boosted Decision Trees ◽  
Learning Event ◽  
Energy Physics

scholarly journals Accelerator, Reactor and Atmospheric Neutrino Data: A Three-Flavor Oscillation Analysis

1997 ◽  
Vol 12 (04) ◽  
pp. 781-799 ◽  
Author(s):  
Srubabati Goswami ◽  
Amitava Raychaudhuri ◽  
Kamales Kar
Keyword(s):  
Experimental Data ◽  
Narrow Range ◽  
Atmospheric Neutrino ◽  
Mass Scale ◽  
Atmospheric Neutrinos ◽  
Mixing Angles ◽  
Oscillation Analysis ◽  
Reactor Neutrinos ◽  
Flavor Oscillation ◽  
Interesting Solution

We perform a three-flavor oscillation analysis of accelerator, reactor and atmospheric neutrino results. Motivated by the experimental data, the mass squared differences are chosen to be Δ12 = Δ13 in the range 0.5–10 eV2 and Δ23 = 10-2 eV2. In such a scenario, the oscillation probabilities for the accelerator and reactor neutrinos involve only two of the mixing angles and one mass scale. But the atmospheric neutrino oscillation is governed by both mass scales and all the three mixing angles. The greater latitude allowed by the scheme leads to some new and interesting solution regions for atmospheric neutrinos in addition to the two-flavor limits usually discussed. However, incorporating the constraints from the previous accelerator searches for neutrino oscillations, a very narrow range of allowed mixing angles survives. This zone is beyond the reach of the on-going accelerator experiments CHORUS and NOMAD with their projected sensitivity.

Sign in / Sign up

Export Citation Format

Share Document