A Unified Framework for Neural Computation and Learning Over Time

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Melacci, Stefano, Betti, Alessandro, Casoni, Michele, Guidi, Tommaso, Tiezzi, Matteo, Gori, Marco
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910610277531648
author Melacci, Stefano
Betti, Alessandro
Casoni, Michele
Guidi, Tommaso
Tiezzi, Matteo
Gori, Marco
author_facet Melacci, Stefano
Betti, Alessandro
Casoni, Michele
Guidi, Tommaso
Tiezzi, Matteo
Gori, Marco
contents This paper proposes Hamiltonian Learning, a novel unified framework for learning with neural networks "over time", i.e., from a possibly infinite stream of data, in an online manner, without having access to future information. Existing works focus on the simplified setting in which the stream has a known finite length or is segmented into smaller sequences, leveraging well-established learning strategies from statistical machine learning. In this paper, the problem of learning over time is rethought from scratch, leveraging tools from optimal control theory, which yield a unifying view of the temporal dynamics of neural computations and learning. Hamiltonian Learning is based on differential equations that: (i) can be integrated without the need of external software solvers; (ii) generalize the well-established notion of gradient-based learning in feed-forward and recurrent networks; (iii) open to novel perspectives. The proposed framework is showcased by experimentally proving how it can recover gradient-based learning, comparing it to out-of-the box optimizers, and describing how it is flexible enough to switch from fully-local to partially/non-local computational schemes, possibly distributed over multiple devices, and BackPropagation without storing activations. Hamiltonian Learning is easy to implement and can help researches approach in a principled and innovative manner the problem of learning over time.
format Preprint
id arxiv_https___arxiv_org_abs_2409_12038
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Unified Framework for Neural Computation and Learning Over Time
Melacci, Stefano
Betti, Alessandro
Casoni, Michele
Guidi, Tommaso
Tiezzi, Matteo
Gori, Marco
Machine Learning
Artificial Intelligence
This paper proposes Hamiltonian Learning, a novel unified framework for learning with neural networks "over time", i.e., from a possibly infinite stream of data, in an online manner, without having access to future information. Existing works focus on the simplified setting in which the stream has a known finite length or is segmented into smaller sequences, leveraging well-established learning strategies from statistical machine learning. In this paper, the problem of learning over time is rethought from scratch, leveraging tools from optimal control theory, which yield a unifying view of the temporal dynamics of neural computations and learning. Hamiltonian Learning is based on differential equations that: (i) can be integrated without the need of external software solvers; (ii) generalize the well-established notion of gradient-based learning in feed-forward and recurrent networks; (iii) open to novel perspectives. The proposed framework is showcased by experimentally proving how it can recover gradient-based learning, comparing it to out-of-the box optimizers, and describing how it is flexible enough to switch from fully-local to partially/non-local computational schemes, possibly distributed over multiple devices, and BackPropagation without storing activations. Hamiltonian Learning is easy to implement and can help researches approach in a principled and innovative manner the problem of learning over time.
title A Unified Framework for Neural Computation and Learning Over Time
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2409.12038