Estimation of high-dimensional change-points under a group sparsity structure

Change-points are a routine feature of ‘big data’ observed in the form high-dimensional data streams. In many such streams, component series possess group structures and it is natural to assume that changes only occur small number all groups. We propose new change point procedure, called groupInspect, exploits sparsity structure estimate projection direction so as aggregate information across successfully change-point mean series. prove estimated minimax optimal, up logarithmic factors, when sizes comparable order. Moreover, our theory provide strong guarantees on rate convergence location estimator. Numerical studies demonstrates competitive performance groupInspect wide range settings real example confirms practical usefulness procedure.