remove repilcated

sections/method.tex

@@ -118,9 +118,6 @@ \subsection{Co-clustering on Small Submatrices}
 \subsubsection{Atom-co-clustering Algorithm}
 Our framework, which encompasses both partitioning and ensembling, offers remarkable flexibility, allowing it to be compatible with a wide range of atom-co-clustering methods. For the sake of making this paper comprehensive and self-contained, we provide an introduction to the atom-co-cluster method herein. The only requirement for an atom-co-clustering method to be compatible with our framework is that it must be able to identify co-clusters under a given criterion with a probability $p$, or more relaxed conditions, has a lower bound estimate of the probability of identifying co-clusters equipped with a validation mechanism.
 
-\subsubsection{Atom-co-clustering Algorithm}
-Our framework's flexibility allows it to be compatible with various atom-co-clustering methods. For an atom-co-clustering method to integrate with our framework, it must identify co-clusters under a given criterion with a certain probability $p$ or provide a lower bound estimate of this probability with a validation mechanism.
-
 \subsubsection{Graph-based Spectral Co-clustering Algorithm}
 Spectral co-clustering (SCC) is widely used for high-dimensional data due to its effectiveness in partitioning data into clusters based on graph Laplacian eigenvectors \cite{vonluxburg2007TutorialSpectralClustering}. SCC constructs a bipartite graph $G=(U,V,E)$ from the data matrix $A$, where $U$ and $V$ represent rows and columns, respectively, and $E$ are weighted edges indicating relationships.