Performance of Pattern Recognition Algorithms in Identifying Banking Networks

A review of literature shows that bank branching can be influenced by economic, geographic, legal, cultural and demographic characteristics apart from institutional decision making. Identifying the branching pattern can help in answering many questions like how a particular banking group is different from other? It can reveal secrets regarding strategic decision making about diversification and growth of banking networks. Present study apply machine learning algorithms to recognize the branching pattern of banking networks in India and compare the differences across groups and types. An evaluation of the performance of the algorithms shows that it is very useful employ a narrative approach for getting important insights from the data compared to traditional approaches. The study also provides a visual map evaluation to enrich the outcomes from the pattern recognition exercise.