Baldi

Bioinformatics: The Machine Learning Approach.

Machine learning approaches play a significant role in bioinformatics due to the abundance of highly variable data and the lack of comprehensive theories. This tutorial provides a broad overview of machine learning approaches in bioinformatics. Specifically, the main topics are:

	   1. The Bayesian probabilistic framework for modeling and induction as
	      the common foundation for all machine learning algorithms.
	   2. A brief presentation of a number of important classes of models
	      routinely used in bioinformatics applications such as: neural
	      networks, hidden Markov models, stochastic grammars, and belief
	      networks and the corresponding learning algorithms.
	   3.  Examples of specific applications, such as:
	       -neural networks for protein functional sites and secondary structure
	        prediction;
	       -hidden Markov models for data base searches, multiple alignments, and
	        pattern discovery;
	       -hidden Markov models for gene finding and promoter prediction;
	       -stochastic grammars and RNA modeling;

Reference: P. Baldi and S. Brunak (1998) "Bioinformatics: the Machine Learning Approach" MIT Press.