Random Prism: An Alternative to Random Forests

Abstract

Ensemble learning techniques generate multiple classifiers, so called base classifiers, whose combined classification results are used in order to increase the overall classification accuracy. In most ensemble classifiers the base classifiers are based on the Top Down Induction of Decision Trees (TDIDT) approach. However, an alternative approach for the induction of rule based classifiers is the Prism family of algorithms. Prism algorithms produce modular classification rules that do not necessarily fit into a decision tree structure. Prism classification rulesets achieve a comparable and sometimes higher classification accuracy compared with decision tree classifiers, if the data is noisy and large. Yet Prism still suffers from overfitting on noisy and large datasets. In practice ensemble techniques tend to reduce the overfitting, however there exists no ensemble learner for modular classification rule inducers such as the Prism family of algorithms. This article describes the first development of an ensemble learner based on the Prism family of algorithms in order to enhance Prism's classification accuracy by reducing overfitting.

An Information-Theoretic Approach for Setting the Optimal Number of Decision Trees in Random Forests

“…It is a type of algorithm that is known as an Ensemble-based classifier, i.e. a class of algorithms where the final classifier is taken as an aggregate of more than one classifier in order to gain better and more accurate results [2]. In the case of RF, intermediate classifiers are modeled as a collection of un-pruned decision trees.…”

“…When the model is in used for classification purposes, the decision tree is traversed starting at the root node and following the appropriate values until a leaf node is reached, giving this node the class label value to the instance (i.e., the node classifies the instance). RF is an example of Ensemble-based Classification method [11], where multiple classifiers are generated and combined results are used to give a final classifier which increases overall accuracy [2].…”

“…Increasing the accuracy and performance of Data Classification (e.g., [7]) in Data Mining and Machine Learning [14] has been a hot topic and even more so in recent years (e.g., [1,2,8]). The Classification process aims at annotating all the instances of a dataset by so-called class labels.…”

Random rules from data streams

“…In [5] the authors present the Random Prism algorithm. Random Prism classifier is inspired from the Prism family of algorithms [2], the Random Decision Forests and Random Forests approaches.…”

Building a Classifier Employing Prism Algorithm with Fuzzy Logic

“…'Separate and conquer' induces a set of IF..THEN rules. However the most notable development using the 'separate and conquer' approach is the Prism family of algorithms [8].…”