Agent-based Learning System

The system learns and makes predictions/suggestions on expert provided input. This software is generic and makes prediction on user-specified input. As the learning advances prediction accuracy improves with each new input.

When the user wants to know which will be the outcome of certain problem, he/she feeds the system with similar problems, specify important characteristics and problem solution. The system learns on user training inputs and outcomes, and then makes a prediction of result on completely new input.

Following is a scenario: Classifying scaffold properties to classes A, B, C or D.

agent1

Intelligent system initialization

Intelligent system output

 

agent2

Experimental results of studied application domains show predication accuracy of more then > 90%.