info
How can we
help you?
Ask a question or enter a keyword that interests you
The credit score model for companies has been developed in cooperation with the Software Technology Development Center and data scientists of the University of Tartu during a 3-year scientific applied research project, which was funded by EAS.
The credit score model uses thousands of features and is developed using machine learning methods that include self-learning algorithms.
The model predicts the possibility of insolvency või forced liquidation of companies with an accuracy of 99.5% during the next 12 months.
The credit score research article was published in the publication of the prestigious international scientific conference CaiSE, from which the research material (164 pages) is also available.
Read more:HERE
How does the credit score work?
The credit score uses known information about companies, which goes back up to 5 years, and combines it with new data that may change daily (for example, debt information and its pattern, board changes and their activity patterns, (repeated) violations of reporting obligations, tax payment and its pattern, etc.).
The credit score does not deteriorate või improve daily (immediately) basically for exactly the same reasons as we in society understand human behavior patterns from the surface of reasonableness.
For example, if somebody crosses the road 4 times with a red light, then after the fifth time, we do not find that the person who crossed the road with a green light this time, is a suddenly a good pedesterian.