Maximum entropy methods for Bayesian analysis in chemistry

Faculty: Science
The project will develop a robust approach to use maximum entropy methods to model likelihood functions for a Bayesian analysis of chemical data. This will allow values and confidence intervals to be obtained from chemical measurements that are better ...

The project will develop a robust approach to use maximum entropy methods to model likelihood functions for a Bayesian analysis of chemical data. This will allow values and confidence intervals to be obtained from chemical measurements that are better (more statistically valid) than those obtained by present methods. The new approach will be applied to important current problems in chemical measurement: the uncertainty of atomic weights of elements, limits on the detection of illegal drugs, and for sports drug testing.

The project will develop a robust approach to use maximum entropy methods to model likelihood functions for a Bayesian analysis of chemical data. This will allow values and confidence intervals to be obtained from chemical measurements that are better (more statistically valid) than those obtained by present methods. The new approach will be applied to important current problems in chemical measurement: the uncertainty of atomic weights of elements, limits on the detection of illegal drugs, and for sports drug testing.

Project team

Key contact

61293854713 or 0411 286 480
b.hibbert@unsw.edu.au