EquiTensors: Learning Fair Integrations of Heterogeneous Urban Data

Neural methods are state-of-the-art for urban prediction problems such as transportation resource demand, accident risk, crowd mobility, and public safety. Model performance can be improved by integrating exogenous features from open data repositories (e.g., weather, housing prices, traffic, etc.), but these uncurated sources are often too noisy, incomplete, and biased to use directly. We propose to learn integrated features from heterogeneous datasets that can be reused across a variety of tasks. We align datasets to a consistent spatio-temporal domain, then describe an unsupervised model based on denoising autoencoders to learn shared features. We extend this core integrative model with adaptive weighting to prevent certain datasets from dominating the signal. To combat discriminatory bias, we use adversarial training to remove correlations with a sensitive attribute (e.g., race or income). Experiments with 23 input datasets and 4 real applications show models using pre-trained features outperform those that ignore exogenous features and are competitive with supervised "oracle" models that use hand-selected datasets while significantly improving fairness. We conclude that our pre-trained integrated features can improve model performance and reduce discriminatory effects for complex prediction applications, broadening the utility of uncurated civic open data repositories.