Encrypted traffic identification has great significance to regulate illegal data, detect network attacks and protect users' information. Here we propose a novel approach to fast identify encrypted traffic via large-scale sparse screening. We investigate randomness features using Lasso regression to select the most relevant features. To make it more efficient to solve large-scale problems, we employ Enhanced Dual Polytope Projections(EDPP) screening rule to remove irrelevant features quickly. The identification is performed with the help of Extreme Learning Machine (ELM) because of its better identification and faster speed. Experimental results show that the method is efficient and effective in encrypted traffic identification.