ﺑﺮآورد وﺟﻪ ﻧﻘﺪ ورودي و ﺧﺮوﺟﻲ ﺷﻌﺐ ﺑﺎﻧﻚ ﺗﺠﺎرت ﺑﺮاي ﻣﺤﺎﺳﺒﺔ وﺟﻪ ﻧﻘﺪ ﻣﻮرد ﻧﻴﺎز ﺷﻌﺒﻪ ﻫﺎ ﺑﺎ اﺳﺘﻔﺎده از ﺗﺤﻠﻴﻞ ﭼﻨﺪﻣﺘﻐﻴﺮة ﺧﻮﺷﻪ ﺑﻨﺪي ﺑﻴﺰي و ﭘﻴﺎده ﺳﺎزي آن در ﺷﺒﻜﻪ ﻫﺎي ﻋﺼﺒﻲ

Estimation of Input & Output Cash of Tejarat Branches in order to Calculate Branches’ Required Cash Via Multivariate Bayesian Clustering Analysis and the Implementation in Neural Network

ﭼﻜﻴﺪه: ﻣﻮﺿﻮع ﻛﻔﺎﻳﺖ وﺟﻪ ﻧﻘﺪ در ﺑﺎﻧﻚ ﻫﺎ، ﻳﻜﻲ از ﻣﺴﺎﺋﻞ ﻣﻬﻢ ﺑﺮاي ﻣـﺪﻳﺮان و ﺑـﻪ ﺧﺼـﻮص رؤﺳﺎي ﻫﺮ ﺷﻌﺒﻪ ﺑﻪ ﺷﻤﺎر ﻣﻲ رود؛ ﭼﺮا ﻛﻪ ﻛﻤﺒـﻮد وﺟـﻪ ﻧﻘـﺪ روزاﻧـﻪ در ﺻـﻨﺪوق ﺷـﻌﺒﻪ ﺑـﻪ ﻋـﺪم ﭘﺎﺳﺨﮕﻮﻳﻲ ﺑﻪ ﻧﻴﺎز ﻣﺸﺘﺮي ﻣﻲ اﻧﺠﺎﻣﺪ و از ﺳﻮي دﻳﮕﺮ، ﻣﺎزاد وﺟﻪ ﻧﻘﺪ در ﺷـﻌﺒﻪ ﻣﻮﺟـﺐ اﻓـﺰاﻳﺶ ﻫﺰﻳﻨﻪ ﺑﺎﺑﺖ اﻧﺘﻘﺎل آن ﺑﻪ ﺧﺰاﻧﺔ ﺑﺎﻧﻚ ﻣﻲ ﺷﻮد. از اﻳﻦ رو ﺑﺎﻧﻚ ﻫﺎ ﻫﻤﻮاره درﺻﺪد ﺗﻌﻴﻴﻦ ﻣﻘﺪار وﺟﻪ ﻧﻘﺪ ﻣﻮرد ﻧﻴﺎز ﺧﻮد ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ ﻋﻤﻠﻴﺎت روزاﻧﻪ ﻫﺴﺘﻨﺪ. ﺑﻪ ﻫﻤﻴﻦ ﻣﻨﻈﻮر در اﻳﻦ ﻣﻘﺎﻟـﻪ، ﺷـﻌﺐ ﺑﺎﻧـﻚ ﺗﺠﺎرت ، ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ ﺗﻨﻮع ﺑﻴﻦ ﺷﻌﺐ، ﺑﺎ دو روش ﺧﻮﺷﻪ ﺑﻨﺪي ﺳﻠﺴﻠﻪ ﻣﺮاﺗﺒﻲ و ﺧﻮﺷﻪ ﺑﻨﺪي ﺑﺮﻣﺒﻨـﺎي روﻳﻜﺮد ﺑﻴﺰي در ﺧﻮﺷﻪ ﻫﺎي ﻣﺘﺸﺎﺑﻪ دﺳﺘﻪ ﺑﻨﺪي ﺷﺪﻧﺪ؛ ﺳﭙﺲ ﺑﺎ در ﻧﻈﺮ ﮔﺮﻓﺘﻦ ﻧﺘﺎﻳﺞ ﺧﻮﺷﻪ ﺑﻨـﺪي ﻣﻘﺪار وﺟﻪ ﻧﻘﺪ ورودي و ﻧﻴﺰ وﺟﻪ ﻧﻘﺪ ﻣﺼﺮﻓﻲ از ﻃﺮﻳﻖ ﺷﺒﻜﻪ ﻫﺎي ﻋﺼﺒﻲ ﺑﺮآورد ﺷﺪ ﺗـﺎ از اﻳـﻦ ﻃﺮﻳﻖ اﻣﻜﺎن ﻣﺤﺎﺳﺒﺔ وﺟﻪ ﻧﻘﺪ ﻻزم ﺑﺮاي ﺷﻌﺐ ﻓﺮاﻫﻢ ﺷﻮد. ﻧﺘﺎﻳﺞ ﺗﺤﻘﻴﻖ ﻧﺸﺎن ﻣﻲ دﻫﺪ، ﺑـﺮآورد وﺟﻪ ﻧﻘﺪ ﻣﺼـﺮﻓﻲ و ورودي ﺷـﻌﺐ ﺑﺎﻧـﻚ ﺑـﺎ اﺳـﺘﻔﺎده از ﺷـﺒﻜﺔ ﻋﺼـﺒﻲ و ﻟﺤـﺎظ ﻛـﺮدن ﻧﺘـﺎ ﻳﺞ ﺧﻮﺷﻪ ﺑﻨﺪي ﺷﻌﺐ ﺑﺎ روﻳﻜﺮد ﺑﻴﺰي، داراي دﻗﺖ ﺑﻴﺸﺘﺮي ﻧﺴﺒﺖ ﺑﻪ ﻧﺘـﺎﻳﺞ ﺧﻮﺷـﻪ ﺑﻨـﺪي ﺷـﻌﺐ ﺑـﺎ روش ﻣﻌﻤﻮل اﺳﺖ.

Cash adequacy in banks’ branches is considered as the significant issues for branch managers; because the daily cash shortage in branches’ funds might lead to the lack of fulfilling customers’ needs. On the other hand, cash surplus in branches will increase the expenses which arise from its transfer to the banks’ treasuries. Therefore, banks have always been attempting to estimate their required cash according to their daily operations and. In this regard, iIn this article, branches of Tejarat Bank, with regard to their diversity, have been classified in similar clusters with the two methods of hierarchical clustering and clustering based on Bayesian approach .Then, based on the results obtained from the clustering, the input cash to the branches as well as the cash consumption in the branches were estimated through the neural networks, which made it possible to calculate the required cash in branches. The results show that the estimation of input and consumed cash of branches using neural network and regarding the results obtained from Bayesian approach for branches clustering enjoys higher precision in comparison to the results obtained from the classic methods of clustering.