Speaker
Description
IceCube DeepCore is an infill of the IceCube Neutrino Observatory designed to study neutrinos with energies as low as 5 GeV. Reconstruction and classification tasks near the lower energy threshold of IceCube DeepCore are especially difficult due to the low number of detected photons per neutrino event. Many neural networks have been developed for these tasks, and there are many ways we could consider training these neural networks to facilitate specific behaviors in the neural networks. In order to take advantage of this variety in architectures and training techniques we could use model stacking techniques with boosted decision trees (BDTs) to combine the outputs of multiple neural networks and improve overall performance. In this talk we show how this technique can improve inelasticity reconstruction with convolutional neural networks (CNNs) trained with different methods and improve neutrino flavor classification using two CNNs with different model architectures.
