• ensemble
• financial distress
• forecast
• machine learning
• nerual networks
• random forest
• svm

In this study we will build a fully automatic workflow of machine learning technique quickly adaptable for similar data on a relateddomain. In short this workflow will start will reshape the time series data, select the most important feature to consider, train different models, ensemble their result and propose a likelihood for predict the future financialdistress of a banks, in order foreseen bankruptcy. The final workflow outputwill also be associated with several intermediate steps that will give a betterunderstanding of the data analysed.