RUUG - Technology
- Minimum Message Length (MML) inference, a machine learning technique based on information theory developed at Monash University, is at the forefront of AI technology. RUUG members are involved in both the further fundamental development of its theory and its application. The data mining tools coming from this work are quite powerful as well as being robust in dealing with noise in the data.
- Bayesian Networks (BNs) are a specific technology which is in growing demand world-wide, as a result of newly developed Bayesian techniques of automating probabilistic reasoning under uncertainty. RUUG has been involved in using BNs in decision support systems and also is at the forefront in efforts to automate the learning of BNs from statistical data.
While RUUG's expertise in MML inference and BNs is world renowned, we also use a variety of other data mining techniques with the focus on finding the most suitable and efficient method for the task at hand.
- genetic algorithms, simulated annealing and Gibbs sampling
- clustering and decision trees
- artificial neural networks
- non-parametric statistics
- support vector machines