Data analytics

Data analytics is a service provided by the University of Newcastle’s Centre for Bioinformatics Biomarker Discovery and Information-Based Medicine (CIBM). It provides scientific, mathematically-literate, supercomputing based processes for examining large raw datasets commonly found in modern society’s deluge of information.

By drawing attention to the relevant characteristics of data it can assist organisations in making better decisions, or in proving or disproving existing theories and assumed working hypotheses. CIBM provides world-class expertise and customisable solutions based on state-of-the-art optimisation techniques linked together with innovative data mining and machine intelligence methodologies.

Facilities and services

CIBM’s state-of-the-art high performance computing facilities amount to an aggregated processing power of approximately 10 Tera FLOPS in 176 cores, 7168 GPU cores, 848 Gb RAM, 10 Tb disk space in two clusters of machines.

  • Darwin computer cluster: with 128 CPUs, is arguably the largest computer entirely dedicated to translational research in Australia.
  • Fisher computer cluster: with a total of sixteen Fermi based GPU cards, combined with the novel software tools being developed, it provides a unique platform in Australia for the study of large scale problems in the area of knowledge discovery from databases.
  • We have access to a team of international specialists that could be recruited to be part of a team.

Data analytics encompasses several activities including:

  • Data Profiling (a mathematically driven analysis of the quality of the data)
  • Exploratory Data Analysis (which helps to find new features and attributes of the data)
  • Confirmatory Data Analysis (a true or false hypotheses test)
  • Identification of associations of interest in the data (selection of characteristics present in the data whose joint presence suggest a higher order characteristic)
  • Path analysis of temporal events (identification of characteristics in time-stamped data)
  • Classification methods
  • Clustering (using state-of-the-art optimisation algorithms to find structures in the data and define mathematical separations between the groups)
  • Predictive Analytics

Key personnel

Professor Pablo Moscato, CIBM Co-Director

Dr. Regina Berretta, CIBM Member, Head of Discipline, Computer Science and Software Engineering

Potential applications

All organisations that store information on their normal operations can use Data Analytics to better understand their operational models, business environment, and to assist in decision making. A comprehensive Data Analytics procedure should be implemented to guarantee the quality of the data and its usefulness for later stage investigation.

Commercialisation and collaboration

CIBM is a solution provider actively pursuing industry engagement for the purpose of contract research and partnerships. CIBM can offer tailored solutions which can be implemented via collaboration, contract research, government sponsored linkage grants that could help business partners to develop a common IP with our institution.

Capabilities PDF: 
X
Loading