| id |
ecaade2020_284 |
| authors |
Tan, Rachel, Patt, Trevor, Koh, Seow Jin and Chen, Edmund |
| year |
2020 |
| title |
Exploration & Validation - Making sense of generated data in large option sets |
| doi |
https://doi.org/10.52842/conf.ecaade.2020.1.653
|
| source |
Werner, L and Koering, D (eds.), Anthropologic: Architecture and Fabrication in the cognitive age - Proceedings of the 38th eCAADe Conference - Volume 1, TU Berlin, Berlin, Germany, 16-18 September 2020, pp. 653-662 |
| summary |
The project is a real-world case study where we advised our client in the selection of a viable and well-performing design from a set of computationally generated options. This process was undertaken while validating the algorithmic generative process and user-defined evaluation criteria through scrutinizing the other alternative options to ensure ample variability was considered. Optimisation algorithms were not ideal as low performing options were not visible to validate variability. We established variability by extracting the different groups of options, proving to the client that various operational behaviours were present and accounted for. In order to sieve through the noise and derive meaningful results, we employed methods to filter through thousands of options, including: k-means clustering, archetypal labelling and analysis, pareto front analysis and visualisation overlays. We present a sense-making and decision-making process that utilizes principles of genetic algorithms and analysis of multi-dimensional user-derived evaluation scores. To enable the client's confidence in the computational model, we proved the effectiveness of the generative model through communicating and visualizing the impact of different criterias. This ensured that operational needs were considered. The visualization methods we employed, including pareto front extraction and analysis eventually helped our clients to arrive at a decision. |
| keywords |
generative design; validation; multi-objective optimisation; k-means; pareto front; decision-making |
| series |
eCAADe |
| email |
|
| full text |
 file.pdf (6,979,371 bytes) |
| references |
Content-type: text/html
Access Temporarily Restricted
Access Temporarily Restricted
Too many requests detected. Please wait 60 seconds or verify that you are a human.
If you are a human user and need immediate access, you can click the button below to continue:
If you continue to experience issues, please open a ticket at
papers.cumincad.org/helpdesk
|
| last changed |
2022/06/07 07:56 |
|
