SANGRIA: Stacked Autoencoder Neural Networks for Indoor Localization
SANGRIA utilizes stacked autoencoder neural networks with gradient boosted trees to address device heterogeneity challenges in indoor localization, resulting in significantly lower average localization errors.