Synthetic Data Generation for Autonomic Computing

Research output: Contribution to conferencePaperpeer-review

Abstract

This paper discusses an approach that integrates data generation capabilities into the Autonomic Computing MAPE-K (Monitor Analyse Plan Execute and Knowledge Loop) to mitigate problems with data scarcity in autonomous space missions. The purpose of this work is to enhance the decision-making abilities of an Autonomic Manager by providing it with the ability to use simulation and data generation. A Conditional Tabular Generative Adversarial Network (CTGAN) is used to generate new synthetic datasets. Synthetic datasets are then evaluated to assess their utility. The evaluation results show that synthetic data can closely resemble the original data. However, this paper does not address the challenges of equipping a swarm with the necessary hardware, focusing instead on the feasibility of the proposed data generation pipeline.
Original languageEnglish
Number of pages7
Publication statusAccepted/In press - 7 Jan 2025
EventThe Twenty First International Conference on Autonomic and Autonomous Systems - Mercure Lisboa Hotel, Lisbon, Portugal
Duration: 9 Mar 202513 Mar 2025
Conference number: 21st
https://www.iaria.org/conferences2025/CfPICAS25.html

Conference

ConferenceThe Twenty First International Conference on Autonomic and Autonomous Systems
Abbreviated titleICAS 2025
Country/TerritoryPortugal
CityLisbon
Period9/03/2513/03/25
Internet address

Keywords

  • Autonomic Computing
  • Generative adversarial network
  • CTGAN
  • MAPE-K

Fingerprint

Dive into the research topics of 'Synthetic Data Generation for Autonomic Computing'. Together they form a unique fingerprint.

Cite this