Florian Rossmannek - researchr alias

researchr

You are not signed in
Sign in
Sign up

Viewing Publication 1 - 6 from 6

2024

Gradient Descent Provably Escapes Saddle Points in the Training of Shallow ReLU NetworksPatrick Cheridito, Arnulf Jentzen, Florian Rossmannek. jota, 203(3):2617-2648, December 2024. [doi]

2023

The curse of dimensionality and gradient-based training of neural networks: shrinking the gap between theory and applicationsFlorian Rossmannek. PhD thesis, ETH Zurich, Zürich, Switzerland, 2023. [doi]

2022

A proof of convergence for gradient descent in the training of artificial neural networks for constant target functionsPatrick Cheridito, Arnulf Jentzen, Adrian Riekert, Florian Rossmannek. jc, 72:101646, 2022. [doi]

Landscape Analysis for Shallow Neural Networks: Complete Classification of Critical Points for Affine Target FunctionsPatrick Cheridito, Arnulf Jentzen, Florian Rossmannek. jns, 32(5):64, 2022. [doi]

Efficient Approximation of High-Dimensional Functions With Neural NetworksPatrick Cheridito, Arnulf Jentzen, Florian Rossmannek. tnn, 33(7):3079-3093, 2022. [doi]

2021

Non-convergence of stochastic gradient descent in the training of deep neural networksPatrick Cheridito, Arnulf Jentzen, Florian Rossmannek. jc, 64:101540, 2021. [doi]

Sign in or sign up to see more results.

runs on WebDSL