Vanneschi, LeonardoSimota, Asimina2018-03-262018-03-262018-02-15http://hdl.handle.net/10362/33280Internship Report presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced AnalyticsNowadays, high market competition requires Banks to focus more at individual customers´ behaviors. Specifically, customers prefer a personal relationship with the finance institution and they want to receive exclusive offers. Thus, a successful cross-sell and up- sell personalized campaign requires to know the individual client interest for the offer. The aim of this project is to create a model, that, is able to identify the probability of a customer to buy a product of the bank. The strategic plan is to run a long-term personalized campaign and the challenge is to create a model which remains accurate during this time. The source datasets consist of 12 dataMarts, which represent a monthly snapshot of the Bank’s dataWarehouse between April 2016 and March 2017. They consist of 191 original variables, which contain personal and transactional information and around 1.400.000 clients each. The selected modeling technique is Artificial Neural Networks and specifically, Multilayer Perceptron running with Back-propagation. The results showed that the model performs well and the business can use it to optimize the profitability. Despite the good results, business must monitor the model´s outputs to check the performance through time.engBankCross-sell and up-sellProbabilitiesArtificial neural networksPerformancemaster thesis201865645