But both faculty and students appreciated the program to be a successful attempt in terms of understanding and appreciation of basic science knowledge in the context of health and disease through an integrated learning program incorporating diverse teaching learning methods.
SVM has proven to be a robust and versatile approach for clinical prediction, as demonstrated by its consistently high performance in head-to-head methodological comparisons of diverse machine learning methods performed with fMRI data [ 29] and other high-dimensional clinical datasets such as proteomics [ 30] and genomics [ 31].
The prediction of eukaryotic protein subcellular localisation has been extensively investigated over the past decade using various machine learning methods and based on many diverse protein characteristics (reviewed in [ 13]).
Moreover, the adaptation of problem-based learning for online delivery offers the opportunity to use new teaching methods and diverse learning foci which, when incorporated into an online educational strategy offer advantages for both the students and those who teach them (Gray and Tobin 2010; Campbell et al. 2008; Sancho et al. 2006; Riffell and Sibley 2005; Garrison and Kanuka 2004).
These programs are diverse in both features used and in machine learning methods.
Both technical and pedagogical innovation should be hallmarks of the best learning environments we can create, and which incorporate a wide variety of pedagogical approaches, learning tools, methods and practices to support students' diverse learning modes.
Supervised learning methods are a useful way to combine predictions from diverse sources.
SVM, introduced by Cortes and Vapnik [21], is a relatively new sub-branch of supervised learning methods, and has been shown to be highly effective for diverse computational biology applications [22].
We are currently developing Bayesian learning methods for transcriptional regulatory networks that can make principled statistical inference from diverse sets of data sources, such as the ones already discussed in this work and other functional data.
However, machine learning methods have shown promising potential to establish the complex QSARs for the data sets with diverse ranges of molecular structures and mechanisms.
