Automation and Computational Intelligence for Road Maintenance and Management

9Models of Performance Measures and Quantification in Automation

9.1 Introduction

In order to evaluate the overall performance of a method or a classifier, a wide range of criteria with varying degrees of sensitivity have been identified to classify methods and decide on selection and application. This chapter is of particular importance due to the importance of evaluating the methods in infrastructure management. In general, these indicators are generally complementary, and the adequacy of one indicator alone does not mean the overall guarantee of the method. The results of evaluation methods of models and algorithms are usually highly dependent on the method of analysis, design, and conditions. The results explain how to extract instructions for extracting information from infrastructure, technologies used, weather conditions when extracting information, processor and speed of analysis, and many other related factors. For these reasons, different results from evaluating automated methods can be expected. In this chapter, the performance evaluation methods and indicators are first presented, followed by a summary of the types of general indicators in infrastructure evaluation. Figure 9.1 shows a general classification of the performance appraisal methods and common indicators.

Assessing the true accuracy and performance of automated infrastructure assessment algorithms is significant in choosing the path of analysis and obtaining practical results. Many indices, such as