Formal Concept Analysis Approximations via Atomic Priming

摘要: Formal Concept Analysis (FCA) looks to decompose a matrix of objects-attributes into set sparse matrices capturing the underlying structure formal context. We propose Rank Reduction (RR) method prime approximate FCAs, namely RRFCA. While many existing FCA algorithms are complete, lectic ordering lattice may not minimize search/decomposition time. Initially, RRFCA decompositions unique or complete; however, good closures with high support is learned quickly, and then, made complete. has its novelty in that we new multiplicative two-stage method. First, describe theoretical foundations underpinning our RR approach. Second, provide representative exemplar, showing how can be implemented. Further experiments demonstrate methods efficient, scalable yield time-savings. resulting lend themselves parallelization.