Learning complex classifiers without search

Project: Research

Project Details

Project Description

Machine learning is a critical enabling technology that underlies many advanced developments in commerce,
industry, governance and science. This project investigates a new paradigm for machine learning that learns
without search. This new paradigm will be used to develop new probabilistic learning techniques that address
limitations of the current state-of-the-art. These novel learning algorithms will create more accurate models
of categorical data, will be robust in the presence of data errors, will support incremental learning and will be
well suited to the types of very large data that are becoming increasingly common. This will allow more
accurate models to be learnt from these common types of data than has previously been possible.
StatusFinished
Effective start/end date4/01/1130/04/14

Funding

  • Australian Research Council (ARC): AUD255,000.00
  • Australian Research Council (ARC)
  • Monash University