Search for: All records

We provide the first known analysis of the "Skip gram with negative sampling" procedure for DeepWalk embeddings of graphs drawn from the stochastic block model. 

We study the problem of selecting the largest among n unknown values x1,…,xn given only a single unbiased estimate yi for each xi. We design strategies that are simultaneously admissible (not uniformly dominated by any other strategy) and also never worse than a given baseline such as uniform random selection. We provide an application to stochastic optimization, specifically to the ubiquitous "iterate averaging" procedure in machine learning. 

We developed new algorithms for graph-augmentation in order to optimize the so-called Kirchoff Index of a graph. This is closely related to well-studied problems such as experiment design in statistics. 

Machine learning (ML) models have emerged as the state-of-the-art approach for wireless applications such as transmitter localization. One of the challenges for ML models, however, is their reliance on the abundance of high quality data. Specifically for localization, a significant challenge is to obtain training data that ''covers'' the entire landscape, in order to ensure a high accuracy for the ML approaches. This issue is compounded when trying to localize multiple transmitters. To address this problem, we introduce a new data augmentation pipeline, termed Physics-informed Augmentation and RF Modeling (PhARMNet), that can combine existing data with a physics-based simulation model, producing a larger dataset with an improved coverage of the landscape of interest. Our results show that PhARMNet offers significant advantages over traditional path loss models. For localization, we demonstrate that augmenting training data with PhARMNet-produced samples improves localization accuracy, particularly in out-of-distribution regions and multi-transmitter settings.