Knowledge discovery in the prediction of bankruptcy

Knowledge discovery in databases (KDD) is the process of discovering interesting knowledge from large amounts of data. However, real-world datasets have problems such as incompleteness, redundancy, inconsistency, noise, etc. All these problems affect the performance of data mining algorithms. Thus, preprocessing techniques are essential in allowing knowledge to be extracted from data. This work presents a real world application of knowledge discovery in databases, with the objective of prediction of bankruptcy. For this task fuzzy classification models based on fuzzy clustering are used, which are developed solely from numerical data. This data set has missing values, extreme values and also presents a much smaller bankruptcy class than the not bankruptcy class, which makes it a challenging problem in the scope of KDD.