Having seen Early, Classical and New Wave nonparametrics, along with partitioningbased classification methods, it is time to examine the most recently emerging class of techniques, here called Alternative methods in parallel with contemporary music. The common feature all these methods have is that they are more abstract. Indeed, the four topics covered here are abstract in different ways. Model-averaging methods usually defy interpretability. Bayesian nonparametrics requires practitioners to think carefully about the space of functions being assumed in order to assign a prior. The relevance vector machine (RVM) a competitor to support vector machines, tries to obtain sparsity by using asymptotic normality; again the interpretability is mostly lost. Hidden Markov models pre-suppose an unseen space to which all the estimates are referred. The ways in which these methods are abstract vary, but it is hard to dispute that the degree of abstraction they require exceeds that of the earlier methods.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2009 Springer-Verlag New York
About this chapter
Cite this chapter
Clarke, B., Fokoué, E., Zhang, H.H. (2009). Alternative Nonparametrics. In: Principles and Theory for Data Mining and Machine Learning. Springer Series in Statistics. Springer, New York, NY. https://doi.org/10.1007/978-0-387-98135-2_6
Download citation
DOI: https://doi.org/10.1007/978-0-387-98135-2_6
Published:
Publisher Name: Springer, New York, NY
Print ISBN: 978-0-387-98134-5
Online ISBN: 978-0-387-98135-2
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)