Videos

To achieve the desired result predisoft uses the effiming score technology.


1 Proportionally Stratified Sample Generator. According to the data input, two samples are generated which keep the relation between good and bad where sample 1 shows the effects learned and sample 2 evaluates them.


2. Correlation analyzer. Analyzes all available variables and compares them with the discriminant (the one that identifies on time and late payers) to determine which follow a pattern. Our technology is capable of automatically sorting through an enormous amount of data that would be impossible to analyze manually.


3. Variable analyzer. Once identified, useful variables are evaluated individually using probability distribution methods and the ones that will be used for the creation of the motor score are identified.


  1. 4.Motor generator. A component that contains all the knowledge on how to identify on time and late payers is created.




Once the motor score is ready, its use is quite simple. All the user must do is send it data which it will evaluate and determine whether a new client will be punctual with his payments.



 
 

Technological Aspects

The analysis and definition of the motor applicable to the credit Score follows a particular organization in which predisoft international uses its own technology developed on a robust mathematical and technological foundation. Effimining score does not only base its results on human resources, but on its top of the line technology.

In essence, effiming score technology uses four crucial processes in the construction of a motor that recognizes whether someone is an on time and late payer when credit is requested.

The customer receives the final product of the sequential and systematic four steps, the DLL with which he/she will determine what kind of a payer an entity is. Predisoft is aware however, of the benefit this technology can provide in risk management and so the software allows for both parties to negotiate the acquisition of the license here in mentioned.



Scoring Analyzer


1. It finds the discriminant variables and defines the modality according to each one. 






DB Manager


2. Builds a random database based on the data provided by the client from which it can learn and another from which it can test.






Discriminant analysis motor


3. Through the learning data base, Factorial Correspondence and Discriminant analysis, a motor is created which identifies an on time or late payer. This process creates a DLL based on learned concepts, which is later used to process new applicants. 






Portfolio Analyzer


4. Using the motor previously generated by the Discriminant Analysis the portfolio is divided into two groups, potentially on time and late payers.