Automation and Computational Intelligence for Road Maintenance and Management

8Classification Methods and Its Applications in Infrastructure Management

8.1 Introduction

Classification is a type of prediction and calculation method in which a method is designed to guess the placement of data in a category, whereby the output is categorized and a class is located. Comparatively, in a regression method, the output is numerical. Descriptive modeling, or clustering, assigns each input to a cluster so that similar inputs are placed in the same cluster. Finally, by observing the relationship between similar data, one can discover rules in the form of association rules about the relationship between inputs.

By common definition, classification is an important application in data science that predicts the attribution of a variable (goal or class) based on the construction of a model using one or more numerical variables (predictors or attributes) and so on. The basis of all variables depends on the categories to which they best belong. The amount of affiliation is calculated by different methods, which fall under four general categories, including:

1. Frequency table
2. Covariance matrix
3. Similarity functions
4. Others

Classification is the most commonly used large dataset classification technique in most research and practical applications of infrastructure management and is widely employed in a variety of applications, such as detection, feature extraction, feature selection, type classification, severity classification, and extent evaluation. In this method, data ...