Fast anchor graph preserving projections

Jikui Wang, Yiwen Wu, Bing Li, Zhenguo Yang, Feiping Nie

Abstract: Highlights•Our model learns a projection matrix while preserving some structure in raw data.•The proposed model has linear time complexity, which makes it adapt to large data sets.•The proposed model has the ability to capture global information of data.•A novel iterative optimization algorithm is proposed to solve the proposed model.•We put forward a theorem to prove the convergence of the proposed algorithm.•Many experiments have verified the efficiency of the proposed FAGPP.