Mining frequent patterns using non-negative matrix factorization

Evolution of Big Data have made giant organizations to embrace data analytics as one of its life line processes. Results from big data analytics have undoubtedly proved that it can bring high productivity in organizations. But several studies also confirms that, there exists many difficulties in handling big data. One of the most acclaimed problem in big data analytics is mining the most frequent patterns that exist in big data. This paper is aimed at proposing a new approach to mine and extract the most frequent patterns using Non-negative matrix factorization (NMF). Results clearly shows that NMF can extract the existing non-obvious patterns that exist in huge volume of data.