Missing Values and Optimal Selection of an Imputation Method and Classification Algorithm to Improve the Accuracy of Ubiquitous Computing Applications
Table 4
Classification algorithms.
Algorithms
Description
C4.5
Estimates the known data using learning rules. C4.5 gradually expands the conditions of the algorithm, splitting the upper node into subnodes using a divide-and-conquer method until it comes to the end node.
SVM
Classifies the unknown class by finding the optimal hyperplane with the maximum margin that reduces the estimation error.
Bayesian network
A probability network with a high posterior probability given the instances. Such a network can provide insight into probabilistic dependencies among the variables in the training dataset.
Logistic classifier
Takes the functional form of logistic CDF (cumulative distribution function). This function relates the probability of some event to attribute variables through regression coefficients and alpha and beta parameters, which are estimated from training data [13].
-nearest neighbor classifier
Simple instance-based learner that uses the class of the nearest training instances for the class of the test instances.
Regression
The class is binarized, and one regression model is built for each class value [14].
