Optimal: Optimization for Machine Learning: From Robustness to Regularization

Abstract

The project has contributed to the development of optimal and efficient algorithms for large scale machine learning and their applications, with results in four main directions: 1) design of algorithms with budgeted space complexity; 2) design of algorithms with minimal time cost; 3) design of algorithms able to exploit data geometry; 4) application to the development of efficient AI systems for humanoid robotics and for for model independent new physics searches. The results of the project have led to new theoretical results, new software and new intelligent systems for robotics. In total 10 academic publications resulted from this grant. Research results have been published and presented in the top venues in the field.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 17, 2022
Accession Number
AD1170158

Entities

People

  • Rosasco Lorenzo

Organizations

  • University of Genoa

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Air Force
  • Air Force Research Laboratories
  • Algorithms
  • Anomaly Detection
  • Artificial Intelligence
  • Banach Space
  • Change Detection
  • Computations
  • Computer Graphics
  • Convergence
  • Errors
  • Estimators
  • Gaussian Processes
  • Graphics
  • High Energy
  • Information Processing
  • Information Science
  • Information Systems
  • Intelligent Systems
  • Learning
  • Machine Learning
  • Neural Networks
  • Optimization
  • Robotics

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Computer Programming and Software Development.
  • Technical Research and Report Writing.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • Autonomy
  • Autonomy - Autonomous System Control
  • Space
  • Space - Spacecraft Maneuvers