An Efficient Algorithm for Finding Continuous Coherent Evolution Bicluster in Time-Series Data

Most traditional biclustering algorithms focus on biclustering model on non-continuous column, which are not suitable for the analysis of the time series gene expression data. We proposes an effective and exact algorithm, which can be used to mine biclusters with coherent evolution on the contiguous columns as well as the complementary biclusters and time-lagged biclusters for the analysis of time series gene expression data. The experimental results show that the algorithm can find biclusters with statistical significance and strong biological relevance. We extend it to the currency data analysis in the financial field and obtain meaningful results.