Machine Learning Basics

• Published in: Intelligent Sensor Networks pp. 3 - 29
• Main Authors: Kapitanova, Krasimira, Son, Sang H.
• Format: Book Chapter
• Language: English
• Published: CRC Press 2013
• Edition: 1
• Subjects: Activity Recognition; Naive Bayes Classifiers; Input Instance; WSN Application; Bayesian Network; CRF Model; Sensor Nodes; HMM; Unsupervised Learning Algorithms; HSMM; Sensor Network Applications; Orienting Subsystem; Dynamic Bayesian Network; Learning Algorithms; Semi-supervised Learning; Attentional Subsystem; Unlabeled Data; Sensor Network; CRF; Conditional Probability Distribution; Static Bayesian Network; Cognitive Wireless Sensor Networks; Machine Learning Algorithms; Van Kasteren
• ISBN: 1138199745; 9781439892817; 1439892814; 9781138199743
• DOI: 10.1201/b14300-2

This chapter presents the most frequently used machine learning algorithms, including clustering, Bayes probabilistic models, Markov models, and decision trees. A major focus of machine learning research is to automatically induce models, such as rules and patterns, from the training data it analyzes. The most frequently used supervised machine learning algorithms include support vector machines, naive Bayes classifiers, decision trees, hidden Markov models, conditional random field, and k-nearest neighbor algorithms. The semi-Markov conditional random fields (SMCRFs) inherits features from both semi-Markov models and Conditional random fields as follows: Hierarchical SMCRFs were used in an activity recognition application on a small laboratory dataset from the domain of video surveillance. In the rest of this section, we describe some of the most commonly used unsupervised learning algorithms. Typical cluster models include the following: We discuss in more detail two of the most common clustering algorithms used in sensor network applications: k-means clustering and density-based spatial clustering for applications with noise clustering.