A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Machine Learning for Spatial and Visual Connectivity
Spatial connectivity measures the distance one needs to traverse to go from any point to any other point in a space. Similarly, visual connectivity measures visual access within a space. Both spatial and visual connectivity analyses require as an input the configuration of the space. To prepare the output that the machine learning (ML) models are expected to predict, both analyses were run using a bespoke simulation engine developed in‐house. A comparative analysis was done between two different ML models' architectures, a convolutional neural network called U‐net and a conditional generative adversarial network called Pix2Pix. The performance of both models using different hyperparameters was compared against a small subset of the generated data set. The best‐performing ones were tested on a bigger subset, a process which continued until the two models were tested with specific hyperparameters on the whole training data set.
Machine Learning for Spatial and Visual Connectivity
Spatial connectivity measures the distance one needs to traverse to go from any point to any other point in a space. Similarly, visual connectivity measures visual access within a space. Both spatial and visual connectivity analyses require as an input the configuration of the space. To prepare the output that the machine learning (ML) models are expected to predict, both analyses were run using a bespoke simulation engine developed in‐house. A comparative analysis was done between two different ML models' architectures, a convolutional neural network called U‐net and a conditional generative adversarial network called Pix2Pix. The performance of both models using different hyperparameters was compared against a small subset of the generated data set. The best‐performing ones were tested on a bigger subset, a process which continued until the two models were tested with specific hyperparameters on the whole training data set.
Machine Learning for Spatial and Visual Connectivity
Carta, Silvio (editor) / Tarabishy,, Sherif (author) / Psarras,, Stamatios (author) / Kosicki, and, Marcin (author) / Tsigkari, Martha (author)
Machine Learning and the City ; 287-291
2022-05-21
5 pages
Article/Chapter (Book)
Electronic Resource
English
Deep learning surrogate models for spatial and visual connectivity
| SAGE Publications | 2020
| Taylor & Francis Verlag | 2024
Did Zipf anticipate spatial connectivity structures?
| Online Contents | 2015
Geometrical and spatial effects on fiber network connectivity
| British Library Online Contents | 2017
Geometrical and spatial effects on fiber network connectivity
| British Library Online Contents | 2017