Global Utilities

Seminars - Abstract

Department of Computer Science & Computer Engineering

Topic:   Pattern Recognition Techniques for Mineral Deposit Prediction from Geoscientific Data
Speaker:   Andrew Skabar
Date:   07-03-2005
Time:   3:00 PM
Venue:   Hooper Lecture Theatre
Abstract:   This seminar will describe research into the development of pattern recognition/machine learning techniques that can be applied to the discovery of mineralization signatures based on multivariate geoscientific data. These signatures can then be used to produce maps depicting areas ranked according to their potential to host deposits of the sought after mineral. The research is motivated by the belief that a successful system could reduce the cost arising out of the risk and uncertainty associated with mineral exploration, and assist in the discovery of ore deposits in areas currently considered non-prospective. The seminar will focus particularly on Bayesian modelling using Markov Chain Monte Carlo (MCMC) techniques applied to multilayer perceptrons (MLPs).
Content Approved by: Head of School
Page maintained by: Applications Programmer
Last Updated: 14 October, 2009

Seminar Information