Shawe-Taylor, J; (2009) Technical perspective: Machine learning for complex predictions. Communications of the ACM , 52 (11) 96 - 96. 10.1145/1592761.1592782.
Full text not available from this repository.
The development of machine learning with respect to its complex predictions is discussed. Machine learning developed during the 1970-1980 with innovations as decision trees, rule-learning methods for expert systems, and self-organizing maps. This line of work not only raised hopes of creating machines that were able to learn, but also understanding the basic mechanisms behind biological learning. The study of support vector machines (SVMs) initiated a new approach to machine learning that focused more on statistical foundations and less on biological plausibility. SVMs were originally formulated for problems in binary classification where the goal is simply to distinguish objects in two different categories. Conventional SVMs can be applied to the problem of part-of-speech tagging by considering each word of a sentence based on the surrounding words.
|Title:||Technical perspective: Machine learning for complex predictions|
|UCL classification:||UCL > School of BEAMS > Faculty of Engineering Science > Computer Science|
Archive Staff Only: edit this record