The bank wants to improve their services. For instance, the bank managers have only vague idea, who is a good client (whom to offer some additional services) and who is a bad client (whom to watch carefully to minimize the bank loses). Fortunately, the bank stores data about their clients, the accounts (transactions within several months), the loans already granted, the credit cards issued. The bank managers hope to improve their understanding of customers and seek specific actions to improve services. A mere application of a discovery tool will not be convincing for them.

To test a data mining approach to help the bank managers, it was decided to address two problems, a descriptive and a predictive one. While the descriptive problem was left open, the predictive problem is the prediction of whether a loan will end successfuly.

The data about the clients and their accounts consist of following relations:

• relation account (4500 objects - each record describes static characteristics of an account,
• relation client (5369 objects) - each record describes characteristics of a client,
• relation disposition (5369 objects) - each record relates together a client with an account i.e. this relation describes the rights of clients to operate accounts,
• relation permanent order (6471 objects) - each record describes characteristics of a payment order,
• relation transaction (1056320 objects) - each record describes one transaction on an account,
• relation loan (682 objects) - each record describes a loan granted for a given account,
• relation credit card (892 objects) - each record describes a credit card issued to an account,
• relation demographic data (77 objects) - each record describes demographic characteristics of a district.

Each account has both static characteristics (e.g. date of creation, address of the branch) given in relation "account" and dynamic characteristics (e.g. payments debited or credited, balances) given in relations "permanent order" and "transaction". Relation "client" describes characteristics of persons who can manipulate with the accounts. One client can have more accounts, more clients can manipulate with single account; clients and accounts are related together in relation "disposition". Relations "loan" and "credit card" describe some services which the bank offers to its clients; more credit cards can be issued to an account, at most one loan can be granted for an account. Relation "demographic data" gives some publicly available information about the districts (e.g. the unemployment rate); additional information about the clients can be deduced from this.

1. Predict if a loan is going to be paid by the client.
2. Predict what loans are more or less risky/beneficial for the bank
3. Group client clusters/types based on their attributes, which can be a helpful tool for future strategies and predictions

1. Make correct predictions about the status goal of the client (if they will or not pay the loan) using the data from past economic cycles.
2. Determine patterns, relations and behaviors of attributes within the dataset provided
3. Determine what attributes are relevant to group clients by, and create clusters based on them that accurately represent probable results from the client base

Quantidade de status para cada tipo(1 e -1): Media da quantidade de loan por tipo de cartão: Media da quantidade de loan por distrito (são muitos distristos portanto os nomes não aparecem todos): Media da quantidade de loan por status (1 e -1): Media do "Salario medio" por status (1 e -1): Media do quantidada final que fica no banco (relation transaction-> balance) por status (1 e -1): Quantidade de account frequency: Quantidade de account frequency por status: Media da quantidade de pagamentos por status: Grafico de correlação de todos os dados