[x] points straight to the place -> point back to the places
[x] This will allow us to study potential sources of ultra-high-energy neutrinos, which are somewhat unknown to this day -> This will allow us to identify and study the to this day enigmatic sources of UHE neutrinos. (potential och somewhat är inte bra i denna mening)
[x] reconstruct the direction of a detected neutrino -> reconstruct the direction of a neutrino from signals in a radio detector
METHOD
[x] Neural networks try to mimic the function of a biological brain, as it can learn by seeing -> A neural network tries to mimic the function of a biological brain, learning by seeing
[x] ARIANNA is not at the South Pole -> remove the ARIANNA
[x] which acts like an eye -> acting like an eye
[x] that the incoming neutrino came from -> where that the incoming neutrino came from.
[x] the space angle error -> the error on the space angle between the simulated and the reconstructed neutrino direction
[x] the direction in space from which the model predicts that the incoming neutrino came from -> the reconstructed direction in space of the incoming neutrino
[x] the space angle error -> the space angle error, i.e. the difference between the incident and reconstructed neutrino direction
RESULTS
[x] the matter ->matter
[x] or hadronic -> and/or hadronic
[x] The forms can be modelled differently, and two common approaches are called Alvarez2009 and ARZ2020. -> Two commonly used approaches to model the radio emission are called Alvarez2009 and ARZ2020.
[x] In the paper -> remove this, unnecessary
CONCLUSIONS
[x] The deep-learning approach to neutrino direction reconstruction has, in our paper, proved itself to be successful -> Our paper proves that deep-learning can be used successfully for neutrino direction reconstruction in radio detectors.
[x] performing well -> its excellent performance
[x] IceCube-Gen2 upgrade -> IceCube-Gen2 (OBS! namnet UPGRADE betyder någonting annat för IceCube)
INTRODUCTION:
METHOD
RESULTS
CONCLUSIONS