The highly complex banking systems that are equipped with remarkable human factor are critical to human mistakes, thus the correction of these incidents are not only of key importance, but it can be very expensive and hardly scalable, especially if we are to utilize human resources excusively.
It is a better choice to replace the human element with automatized processes wherever possible as it makes the data-processing systems easier to supervise, control and maintain. A good solution for these automatized processes would be treating them with Artificial Intelligence, whose related basics I’m going to cover in detail during the course of this essay. I will also illustrate the practical aspects of utilizing these technologies with the help of an example.
Of the bond market instruments, I will also touch on the technical aspects of the representation, the conversion into processable format, the processing itself, the editing and finally the conversion from the format which we were working on to its original standard format. For this I’ll be using the OO-based JAVA language’s reflective qualities and the JAXB API.
I will explain how the core genetic searching algorithm works and then touch upon how many ways can the user – approximately – corrupt bond data, and for each type of error, how much resource does the genetic algorithm need for having the bond repaired, and finally what are the extreme cases where the algorithm can repair a bond within a reasonable timeframe.
In the end I’m concluding what I’ve learned working on this problem and come up with a possible future direction which I think would worth following for making the corrective algorithm more efficient.