1: \begin{abstract}

2: 		This paper introduces a novel framework called \underline{Mo}de-wise \underline{P}rincipal S\underline{u}bspace \underline{P}ursuit (\texttt{MOP-UP}) to extract hidden variations in both the row and column dimensions for matrix data. To enhance the understanding of the framework, we introduce a class of matrix-variate spiked covariance models that serve as inspiration for the development of the \texttt{MOP-UP} algorithm. The \texttt{MOP-UP} algorithm consists of two steps: Average Subspace Capture (\texttt{ASC}) and Alternating Projection (\texttt{AP}). These steps are specifically designed to capture the row-wise and column-wise dimension-reduced subspaces which contain the most informative features of the data. \texttt{ASC} utilizes a novel average projection operator as initialization and achieves exact recovery in the noiseless setting. We analyze the convergence and non-asymptotic error bounds of \texttt{MOP-UP}, introducing a blockwise matrix eigenvalue perturbation bound that proves the desired bound, where classic perturbation bounds fail. The effectiveness and practical merits of the proposed framework are demonstrated through experiments on both simulated and real datasets. Lastly, we discuss generalizations of our approach to higher-order data.

3: 	\end{abstract}

4: