|
Abstract:
|
Banking branch performance evaluation plays a key role in reinvigorating branch networks in competitive marketplace. As a crucial step, the successful branch patterns need to be discovered from diverse perspectives such as capacity, commercial and operational manners, geographic and demographic characteristics. However, current research and industrial efforts have mainly focused on internal patterns. In this paper, we adopt the spatial data mining technique to discover useful patterns of best branches. Data Envelopment Analysis is firstly carried out to define the best branches. Spatial data mining techniques such as clustering analysis are then employed to find out their environmental characteristics. In one real case, those successful patterns incorporating both geographic and demographic characteristics are discovered with high potential to guide real practices.
|
