|
Home
|
Scorto™ Credit Decision Review
Scorto Credit Decision is a user-friendly yet powerful and feature-rich tool to develop credit models. It allows the user to develop credit models using traditional (i.e. logistic regression) as well as state-of-the-art techniques (i.e. neural networks, decision trees).
Another interesting feature of the software is that it offers the possibility to develop models on the basis of expert knowledge rather than statistical analysis. The user has also a wide range of options to explore the data prior to the development of the model, such as dependencies, distribution, histograms, Kohonen maps and statistics, as well as to handle missing values and outliers. A number of tools are also available for the evaluation of the developed models. More precisely, ROC curves and Gini values accompanied by indicators such as Lorenz curves and statistic, and Kolmogorov-Smirnov curve, all help in the development of efficient credit models. Profit and Loss graphs are also available, illustrating the impact of alternative cut-off points on the rates of approved and "bad" requests. The software is accompanied by a very good manual that guides the user step-by-step, making the development of credit models with Scorto Credit Decision even easier.
Coming from an academic background, I have so far relied on statistical/econometric and mathematical software applications for the development of credit models. I believe that Scorto Credit Decision has more options than non-specialized (in credit) statistical/econometric software applications while being more user friendly than non-specialized mathematical software programs that can be very powerful but require advanced knowledge of programming and the techniques that are being applied for the development of the models.
|
