no code implementations • 5 Dec 2023 • Céline Comte, Matthieu Jonckheere, Jaron Sanders, Albert Senen-Cerda
As a second contribution, we show that, under appropriate assumptions, the policy under a SAGE-based policy-gradient method has a large probability of converging to an optimal policy, provided that it starts sufficiently close to it, even with a nonconvex objective function and multiple maximizers.
no code implementations • 11 Aug 2023 • Gianluca Kosmella, Ripalta Stabile, Jaron Sanders
Specifically, we investigate a probabilistic framework for the first design that establishes that the design is correct, i. e., for any feed-forward NN with Lipschitz continuous activation functions, an ONN can be constructed that produces output arbitrarily close to the original.
no code implementations • 4 Oct 2022 • Alexander Van Werde, Albert Senen-Cerda, Gianluca Kosmella, Jaron Sanders
We address this issue and investigate the suitability of these clustering algorithms in exploratory data analysis of real-world sequential data.
no code implementations • 18 Dec 2020 • Oxana A. Manita, Mark A. Peletier, Jacobus W. Portegies, Jaron Sanders, Albert Senen-Cerda
The first theorem applies to dropout networks in the random mode.
no code implementations • 1 Dec 2020 • Albert Senen-Cerda, Jaron Sanders
We analyze the convergence rate of gradient flows on objective functions induced by Dropout and Dropconnect, when applying them to shallow linear Neural Networks (NNs) - which can also be viewed as doing matrix factorization using a particular regularizer.
no code implementations • 6 Feb 2020 • Albert Senen-Cerda, Jaron Sanders
We investigate the convergence and convergence rate of stochastic training algorithms for Neural Networks (NNs) that have been inspired by Dropout (Hinton et al., 2012).