Machine learning

Vishnu Nath, Stephen E. Levinson

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Whenever a problem seems extremely open ended with a large variety of random variables that have an effect on the process, it is impossible for a human programmer to be able to account for every single case. The number of cases increases dramatically with an additional parameter. In such scenarios, probabilistic algorithms have the greatest applicability. In such scenarios, the algorithms need to be given a couple of examples of scenarios it might come across and the algorithm would be able to handle a new scenario with reasonable accuracy. The key word in the previous statement is ‘reasonable’. There is no probabilistic algorithm that will always return the optimum result with a probability of 1. That would make it a deterministic algorithm which, as has just been discussed, cannot handle every potential case. In this chapter, we discuss the algorithms that were employed to successfully complete the experiment.

Original languageEnglish (US)
Title of host publicationSpringerBriefs in Computer Science
PublisherSpringer
Pages39-44
Number of pages6
Edition9783319056050
DOIs
StatePublished - 2014

Publication series

NameSpringerBriefs in Computer Science
Number9783319056050
Volume0
ISSN (Print)2191-5768
ISSN (Electronic)2191-5776

Keywords

  • High Reward
  • Label Data
  • Reinforcement Learning Algorithm
  • Unlabeled Data
  • Unsupervised Learning

ASJC Scopus subject areas

  • General Computer Science

Fingerprint

Dive into the research topics of 'Machine learning'. Together they form a unique fingerprint.

Cite this