1: \begin{abstract}

2: Non-negative least squares~(NNLS)  is one of the most fundamental problems in numeric analysis. It has been widely used in scientific computation and data modeling. 

3: When applying to huge and complex data in big data analytics, the limitations of NNLS algorithm speed and accuracy remain as key challenges. In this paper, we develop a fast and robust anti-lopsided algorithm for NNLS with high accuracy that is totally based on the first order methods. 

4: The main idea of the proposed algorithm is to transform the original NNLS into an equivalent non-negative quadratic programming, which significantly reduces the scaling problem of variables to discover more effective gradient directions.  The proposed algorithm reaches high accuracy and fast speed  with linear convergence $(1 - \frac{1}{2||Q||_2})^k$ in the subspace of passive variables where $\sqrt{n} \leq ||Q||_2 \leq n$, and $n$ is the dimension size of solutions. The experiments on large matrices clearly show the high performance of the proposed algorithm in comparison to the state-of-the-art algorithms.

5: 

6: \vspace{0.5cm}

7: \textbf{Keywords}: Non-negative least squares, Large-scale anti-lopsided algorithm, First order methods.

8: \end{abstract}

9: