Reconciling Classical and Modern (Deep) Machine Learning for Real-World Applications

APHELEIA aims to create robust, interpretable, and efficient machine learning models that require less data by integrating classical methods with modern deep learning, fostering interdisciplinary collaboration.

Subsidie
€ 1.999.375
2023

Projectdetails

Introduction

Despite the undeniable success of machine learning in addressing a wide variety of technological and scientific challenges, the current trend of training predictive models with an ever-growing number of parameters from an ever-growing amount of data is not sustainable.

Challenges of Current Models

These huge models, often engineered by large corporations benefiting from substantial computational resources, typically require learning a billion or more parameters. They have proven to be very effective in solving prediction tasks in various fields such as:

  1. Computer vision
  2. Natural language processing
  3. Computational biology

However, they mostly remain black boxes that are hard to interpret, computationally demanding, and not robust to small data perturbations.

Objectives of APHELEIA

With a strong emphasis on visual modeling, the grand challenge of APHELEIA is to develop a new generation of machine learning models that are:

  • More robust
  • Interpretable
  • Efficient
  • Not reliant on massive amounts of data to produce accurate predictions

To achieve this objective, we will foster new interactions between:

  • Classical signal processing
  • Statistics
  • Optimization
  • Modern deep learning

Methodology

Our goal is to reduce the need for massive data by enabling scientists and engineers to design trainable machine learning models that:

  • Directly encode a priori knowledge of the task semantics and data formation process
  • Automatically prefer simple and stable solutions over complex ones

These models will be built on solid theoretical foundations with convergence and robustness guarantees, which are important for making trustworthy predictions in real-life scenarios.

Implementation and Impact

We will implement these ideas in an open-source software toolbox readily applicable to visual recognition and inverse imaging problems, which will also handle other modalities. This initiative will stimulate interdisciplinary collaborations, with the potential to be a game changer in the way scientists and engineers design machine learning problems.

Financiële details & Tijdlijn

Financiële details

Subsidiebedrag€ 1.999.375
Totale projectbegroting€ 1.999.375

Tijdlijn

Startdatum1-9-2023
Einddatum31-8-2028
Subsidiejaar2023

Partners & Locaties

Projectpartners

  • INSTITUT NATIONAL DE RECHERCHE EN INFORMATIQUE ET AUTOMATIQUEpenvoerder

Land(en)

France

Vergelijkbare projecten binnen European Research Council

ERC Starting...

Cascade Processes for Sparse Machine Learning

The project aims to democratize deep learning by developing small-scale, resource-efficient models that enhance fairness, robustness, and interpretability through innovative techniques from statistical physics and network science.

€ 1.499.285
ERC Starting...

Scalable Learning for Reproducibility in High-Dimensional Biomedical Signal Processing: A Robust Data Science Framework

ScReeningData aims to develop a scalable learning framework to enhance statistical robustness and reproducibility in high-dimensional data analysis, reducing false positives across scientific domains.

€ 1.500.000
ERC Starting...

AI-based Learning for Physical Simulation

This project aims to enhance physical simulations by integrating machine learning with equation-based modeling for improved generalization and intelligibility, applicable across scientific disciplines and engineering.

€ 1.315.000
ERC Advanced...

Control for Deep and Federated Learning

CoDeFeL aims to enhance machine learning methods through control theory, developing efficient ResNet architectures and federated learning techniques for applications in digital medicine and recommendations.

€ 2.499.224
ERC Starting...

Optimizing for Generalization in Machine Learning

This project aims to unravel the mystery of generalization in machine learning by developing novel optimization algorithms to enhance the reliability and applicability of ML in critical domains.

€ 1.494.375