CAGEN - Context-Action Generation for Testing Self-learning Functions
| Autor: | Eric Sax, Marco Stang, Maria Guinea Marquez |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
business.industry
Test data generation Computer science media_common.quotation_subject Context (language use) Machine learning computer.software_genre Test (assessment) Data set Action (philosophy) Mediation Global Positioning System Artificial intelligence business Function (engineering) computer media_common |
| Zdroj: | Advances in Intelligent Systems and Computing ISBN: 9783030732707 IHIET (Strasbourg) |
| Popis: | This paper presents the general concept and the prototypical implementation for generating data for testing self-learning functions. The concept offers the possibility to create a data set consisting of several different data providers. For this purpose, a mediation pattern adapted to the data generation was developed. The concept was applied to test a self-learning comfort function by classifying the context data into three subgroups: real-, sensor-, and user data. This separation allows a more realistic simulation using the data to test a self-learning comfort function and detect possible malfunctions. The CAGEN concept was implemented as a prototype by simulating GPS and temperature data. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|