Publications

Journals

Estimating the Order of an Autoregressive Model Using Normalized Maximum Likelihood, D. F. Schmidt and E. Makalic, IEEE Transactions on Signal Processing, Vol. 59, No. 2, pp. 479-487, 2011.

Fast Computation of the Kullback-Leibler Divergence and Exact Fisher Information Matrix for the First-Order Moving Average Model, E. Makalic and D. F. Schmidt, IEEE Signal Processing Letters, Vol. 17, No. 4, pp. 391-393, 2010.

Universal Models for the Exponential Distribution, D. F. Schmidt and E. Makalic, IEEE Transactions on Information Theory, Vol. 55, No. 7, pp. 3087-3090, 2009.

Minimum Message Length Shrinkage Estimation, E. Makalic and D. F. Schmidt, Statistics & Probability Letters, Vol. 79, No. 9, pp. 1155-1161, 2009.

Conferences and Workshops

MDL Multiple Hypothesis Testing, E. Makalic and D. F. Schmidt, The Fourth Workshop on Information Theoretic Methods in Science and Engineering (WITMSE), August 2011.

The behaviour of the Akaike Information Criterion when applied to non-nested sequences of models, D. F. Schmidt and E. Makalic, Proceedings of the 23rd Australasian Joint Conference on Artificial Intelligence (AI-10), Adelaide, Australia, pp. 223-232, 2010 [best paper].

Review of modern logistic regression methods with application to small and medium sample size problems, E. Makalic and D. F. Schmidt, Proceedings of the 23rd Australasian Joint Conference on Artificial Intelligence (AI-10), Adelaide, Australia, pp. 213-222, 2010.

MML Invariant Linear Regression, D. F. Schmidt and E. Makalic, Proceedings of the 22nd Australasian Joint Conference on Artificial Intelligence (AI-09), Melbourne, Australia, pp, 312-321, 2009. Note, this version of the paper fixes a minor error in equation (20). [code]

Presentations and Seminars

A tutorial on Point Estimation, E. Makalic and D. F. Schmidt, The University of Melbourne, August 2012.

Learning the structure of your data using clustering/mixture modelling, D. F. Schmidt and E. Makalic, The University of Melbourne, August 2012.

Logistic Regression with the Nonnegative Garrote, E. Makalic and D. F. Schmidt, 24th Australasian Joint Conference on Artificial Intelligence, December 2011.

Minimum Message Length Analysis of the Behrens-Fisher Problem, E. Makalic and L. Allison, Solomonoff 85th Memorial Conference, November 2011.

MMLD Inference of Multilayer Perceptrons, E. Makalic and D. F. Schmidt, Solomonoff 85th Memorial Conference, November 2011.

MDL Multiple Hypothesis Testing, E. Makalic and D. F. Schmidt, The Fourth Workshop on Information Theoretic Methods in Science and Engineering (WITMSE), August 2011.

Tutorial: Akaike’s Information Criterion, D. F. Schmidt and E. Makalic, Monash University, December 2008.

Review of modern logistic regression methods, E. Makalic and D. F. Schmidt, The University of Melbourne, October 2010.

Review of modern logistic regression methods with application to small and medium sample size problems, E. Makalic and D. F. Schmidt, 23rd Australasian Joint Conference on Artificial Intelligence, December 2010.

The behaviour of the Akaike Information Criterion when applied to non-nested sequences of models, D. F. Schmidt and E. Makalic, 23rd Australasian Joint Conference on Artificial Intelligence, December 2010.

MML Logistic Regression with Translation and Rotation Invariant Priors, E. Makalic and D. F. Schmidt, 25th Australasian Joint Conference on Artificial Intelligence, December 2012.

Posters

Some machine learning approaches to discovering information in mammograms, D. F. Schmidt, E. Makalic, J. Stone, Ruth M. L. Warren and J. L. Hopper, 2011.

A Minimum Encoding Approach to Analysing GWAS Data, E. Makalic, D. F. Schmidt, M. Jenkins and J. L. Hopper, International Genetic Epidemiology Society (IGES), Hawaii, USA, 2009.

Technical Reports

MMLD Inference of the Poisson and Geometric Models, D. F. Schmidt and E. Makalic, Clayton School of Information Technology, Monash University, 12 pp. Technical report 2008/220.

Shrinkage and Denoising by Minimum Message Length, D. F. Schmidt and E. Makalic, Clayton School of Information Technology, Monash University, 6 pp. Technical report 2008/230.

Efficient Linear Regression by Minimum Message Length, E. Makalic and D. F. Schmidt, Clayton School of Information Technology, Melbourne, 20pp. Technical report 2006/201.

Unpublished Manuscripts

Fisher information in Single Factor Analysis, E. Makalic, The University of Melbourne, 5pp, 2012.