Examining and Evaluating Classification Algorithms Based on Decision Trees

Abstract

Mehr Ali Qasimi

Machine learning learns everything from the data we provide it, uses that data to predict future outcomes, and more. Machine Learning is the process by which computer work more accurately as it learns from the given data. The adoption of machine learning techniques is beneficial in all fields of research. There are many types of machine learning include unsupervised learning, enforcement learning, and supervised learning. Classification is a part or type of supervised machine learning technique. These algorithms are used to identify and predict data in a variety of domains, including replacement statistical methods, search engine fields, and fields with medical certifications. A supervised learning technique called classification assigns a data item to one of several pre-established classifications. One of the most popular approaches for representing classifiers in data classification is the use of decision tree classifiers. A decision tree classification is a machine learning technique that predicts or determines the classes of future data sets when the class labels are unknown by using the predetermined labels from previous known sets. Decision tree classifiers have been suggested for usage in many different disciplines, including medical disease analysis, text categorization, user smartphone classification, pictures, and many more. Numerous decision tree algorithms exist, and they are categorized according to how accurate and costly they are to use.

PDF