muPRL: A Mutation Testing Pipeline for Deep Reinforcement Learning based on Real Faults
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916371828310016 |
|---|---|
| author | Thomas, Deepak-George Biagiola, Matteo Humbatova, Nargiz Wardat, Mohammad Jahangirova, Gunel Rajan, Hridesh Tonella, Paolo |
| author_facet | Thomas, Deepak-George Biagiola, Matteo Humbatova, Nargiz Wardat, Mohammad Jahangirova, Gunel Rajan, Hridesh Tonella, Paolo |
| contents | Reinforcement Learning (RL) is increasingly adopted to train agents that can deal with complex sequential tasks, such as driving an autonomous vehicle or controlling a humanoid robot. Correspondingly, novel approaches are needed to ensure that RL agents have been tested adequately before going to production. Among them, mutation testing is quite promising, especially under the assumption that the injected faults (mutations) mimic the real ones.
In this paper, we first describe a taxonomy of real RL faults obtained by repository mining. Then, we present the mutation operators derived from such real faults and implemented in the tool muPRL. Finally, we discuss the experimental results, showing that muPRL is effective at discriminating strong from weak test generators, hence providing useful feedback to developers about the adequacy of the generated test scenarios. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_15150 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | muPRL: A Mutation Testing Pipeline for Deep Reinforcement Learning based on Real Faults Thomas, Deepak-George Biagiola, Matteo Humbatova, Nargiz Wardat, Mohammad Jahangirova, Gunel Rajan, Hridesh Tonella, Paolo Software Engineering Reinforcement Learning (RL) is increasingly adopted to train agents that can deal with complex sequential tasks, such as driving an autonomous vehicle or controlling a humanoid robot. Correspondingly, novel approaches are needed to ensure that RL agents have been tested adequately before going to production. Among them, mutation testing is quite promising, especially under the assumption that the injected faults (mutations) mimic the real ones. In this paper, we first describe a taxonomy of real RL faults obtained by repository mining. Then, we present the mutation operators derived from such real faults and implemented in the tool muPRL. Finally, we discuss the experimental results, showing that muPRL is effective at discriminating strong from weak test generators, hence providing useful feedback to developers about the adequacy of the generated test scenarios. |
| title | muPRL: A Mutation Testing Pipeline for Deep Reinforcement Learning based on Real Faults |
| topic | Software Engineering |
| url | https://arxiv.org/abs/2408.15150 |