LogSpecT: Feasible Graph Learning Model from Stationary Signals with Recovery Guarantees

Neural Information Processing Systems 

Graph learning from signals is a core task in graph signal processing (GSP).