Hola,
¿eres nuevo aquí?

Registrate y conecta tu empresa con financiación pública, partners y proyectos.

DYNASTY

Financiado

Dynamics-Aware Theory of Deep Learning

The recent advances in deep learning (DL) have transformed many scientific domains and have had major impacts on industry and society. Despite their success, DL methods do not obey most of the wisdoms of statistical learning theor... ver más

30/09/2027

INRIA

1M€

Presupuesto del proyecto: 1M€

Líder del proyecto

INSTITUT NATIONAL DE RECHERCHE EN INFORMATIQU... No se ha especificado una descripción o un objeto social para esta compañía.

TRL 4-5

Financiación concedida El organismo HORIZON EUROPE notifico la concesión del proyecto el día 2022-06-03

Línea de financiación objetivo El proyecto se financió a través de la siguiente ayuda:

ERC-2021-STG: ERC STARTING GRANTS

I+D

Cerrada hace 3 años

0% 100% 100%

1 Participantes

INRIA

1.50M€ | Lider

Proyectos interesantes

TUCLA Theoretical Understanding of Classic Learning Algorithms 2M€ Cerrado

Understanding DL Understanding Deep Learning 1M€ Cerrado

BeyondBlackbox Data Driven Methods for Modelling and Optimizing the Empiric... 1M€ Cerrado

EEBB-I-12-05083 APRENDIZAJE CAUSAL: TEORIA, METODO Y APLICACIONES. 5K€ Cerrado

PrSc-HDBayLe Provable Scalability for high-dimensional Bayesian Learning 1M€ Cerrado

PID2021-127311NB-I00 BASES FISICAS DE LA INTELIGENCIA NATURAL: DE CELULAS A ORGAN... 145K€ Cerrado

Últimas noticias

29-11-2024: IDAE En las últimas 48 horas el Organismo IDAE ha otorgado 4 concesiones

29-11-2024: ECE En las últimas 48 horas el Organismo ECE ha otorgado 2 concesiones

29-11-2024: CMVAMADRID En las últimas 48 horas el Organismo CMVAMADRID ha otorgado 37 concesiones

Duración del proyecto: 63 meses Fecha Inicio: 2022-06-03
Fecha Fin: 2027-09-30

Presupuesto del proyecto 1M€

Descripción del proyecto The recent advances in deep learning (DL) have transformed many scientific domains and have had major impacts on industry and society. Despite their success, DL methods do not obey most of the wisdoms of statistical learning theory, and the vast majority of the current DL techniques mainly stand as poorly understood black-box algorithms. Even though DL theory has been a very active research field in the past few years, there is a significant gap between the current theory and practice: (i) the current theory often becomes vacuous for models with large number of parameters (which is typical in DL), and (ii) it cannot capture the interaction between data, architecture, training algorithm and its hyper-parameters, which can have drastic effects on the overall performance. Due to this lack of theoretical understanding, designing new DL systems has been dominantly performed by ad-hoc, 'trial-and-error' approaches. The main objective of this proposal is to develop a mathematically sound and practically relevant theory for DL, which will ultimately serve as the basis of a software library that provides practical tools for DL practitioners. In particular, (i) we will develop error bounds that closely reflect the true empirical performance, by explicitly incorporating the dynamics aspect of training, (ii) we will develop new model selection, training, and compression algorithms with reduced time/memory/storage complexity, by exploiting the developed theory. To achieve the expected breakthroughs, we will develop a novel theoretical framework, which will enable tight analysis of learning algorithms in the lens of dynamical systems theory. The outcomes will help relieve DL from being a black-box system and avoid the heuristic design process. We will produce comprehensive open-source software tools adapted to all popular DL libraries, and test the developed algorithms on a wide range of real applications arising in computer vision, audio/music/natural language processing.

Conecta tu I+D

Entra hoy

Financiación

Empresas

CTIs/Universidades

Proyectos

Investigadores