Knowledge Engineering and Discovery Research Institute (KEDRI)

Knowledge Engineering and Discovery Research Institute (KEDRI)

Description

KEDRI develops novel information processing methods, technologies and their applications to enhance discoveries across different areas of science and engineering. The methods are mainly based on principles such as brain information processing, evolution, genetics, and quantum physics. KEDRI's six research centres reflect the projects that it undertakes: - novel methods of computational intelligence - neurocomputation and neuroinformatics - bioinformatics - data mining and decision support systems - adaptive pattern recognition - study of creativity

Expertise
  • Information Sciences
  • Artificial Intelligence (Spiking Neural Networks, Fuzzy Systems, Evolutionary Computation)
  • Machine learning
  • Bioinformatics
  • Brain–like Computing and Neuroinformatics
  • Signal, Speech and Image Processing
Research Focus

Two main research focuses of KEDRI are Spiking Neural Network (SNN) and Personalised Modelling (PM) for data analysis.

KEDRI is working on the development of computational methods using Spiking Neural Networks in several research projects for image, video and audio recognition. The applications of these methods include object detection, face recognition, human action recognition, voice identification, fMRI analysis, EEG signal analysis, robot based rehabilitation.

Another main research in KEDRI is developing novel systems and models based on Personalised Modelling for different applications, such as medical decision support system, prediction system for environmental risk, etc.

Unique Capability Proposition

Strong connections with private and public sector organisations from New Zealand and overseas. http://www.kedri.aut.ac.nz/partnerships-and-associations for the full list.

Past Successes
  • 2012/2014, 'Methods of computational intelligence and software–hardware systems for personalised mental health risk analysis, outcome prognosis and rehabilitation', with China Academy of Science (CAS), NZ$300,000
  • 2008/2010, NiCT, Tokyo, Japan, Fast algorithms for cyber–security data stream on–line modelling and analysis, NZ$120,000 p.a.
  • 2002/2007, NERF, Connectionist–Based Intelligent Information Systems, $360,000 p.a
  • 2006, Fonterra, Predicting milk volume production, NZ$78,000
  • 2006, Telecom/Lucent/Medialab, Close loop optimisation, NZ$70,000