Formulation and research of new fitness function in the genetic algorithm for maximum code coverage
| Autor: | K.E. Serdyukov, Z.B. Tsydenov, Tatiana Avdeenko |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Fitness function
Computer science White-box testing Code coverage 020206 networking & telecommunications 02 engineering and technology Path (graph theory) Genetic algorithm 0202 electrical engineering electronic engineering information engineering General Earth and Planetary Sciences Graph (abstract data type) Control flow graph 020201 artificial intelligence & image processing Algorithm General Environmental Science Test data |
| Zdroj: | Procedia Computer Science. 186:713-720 |
| ISSN: | 1877-0509 |
| Popis: | In present paper we investigate an approach to intelligent support of the software white box testing process based on evolutionary paradigm. As part of this approach we solve the urgent problem of finding the optimal set of test data that provides maximum coverage of the code when it is used in the testing process. Traditionally, to solve this problem, an approach is used when the genetic algorithm is used to find the most adapted chromosome, which is a set of test data that ensures passage along the most complex (long) path in the code graph. Many data sets that provide maximum code coverage can be found by repeating this procedure multiple times with preliminary zeroing of the code operation weights corresponding to the chromosomes found earlier. Thus, the fitness function of the genetic algorithm has a simple form, but the process of finding all the data sets is quite long and non-optimal. In this paper, we study a different approach, when an additional component is initially included in the fitness function, which is responsible for the greatest possible difference in the paths of the population, along with a component responsible for the complexity of each path. The basic version of the algorithm is examined and it’s more effective modifications have been proposed to achieve the final goal of obtaining the optimal set of test data covering the flow graph in the least expensive way. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|