Hierarchical Learning-Guided Automatic Model Discovery

Abstract

While participating in the DARPA Data Driven Discovery (D3M) program, we have explored many aspects necessary for automatic model and pipeline discovery. Based on insights gained we built a modular AutoML system performing well on a wide range of task and data types: tabular, image, graph, audio, video, text, etc. We have contributed code used by other participants in the program to enable collaboration and interoperability.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 2021
Accession Number
AD1154910

Entities

People

  • Dawn Song

Organizations

  • University of California, Berkeley

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Air Force
  • Air Force Research Laboratories
  • Algorithms
  • Artificial Intelligence
  • Artificial Intelligence Computing
  • Artificial Intelligence Software
  • Computer Languages
  • Convolutional Neural Networks
  • Data Mining
  • Deep Learning
  • Dimensionality Reduction
  • Information Science
  • Information Systems
  • Knowledge Management
  • Machine Learning
  • Neural Networks
  • Ontologies
  • Reinforcement Learning
  • United States

Fields of Study

  • Computer science

Readers

  • Business Analytics
  • Distributed Systems and Data Platform Development
  • Neural Network Machine Learning.