Augmenting a Deferred Bayesian Network with a Genetic Algorithm to Correct Missed RFID Readings
Author(s)
Darcy, Peter
Stantic, Bela
Sattar, Abdul
Year published
2009
Metadata
Show full item recordAbstract
The alluring promise of a system that would effortlessly identify a mass of objects has been the goal of all research towards Radio Frequency Identification. Unfortunately, in reality, several problems such as missed readings hinder RFID from being as globally utilised as it should be. In this investigation, we present a /emph{Deferred Bayesian Network} solution that has obtained its network configuration through means of /emph{Genetic Algorithm} training. From this research, we have found the optimal amount of chromosomes required to ensure the highest possible result. Additionally, we have compared our methodology ...
View more >The alluring promise of a system that would effortlessly identify a mass of objects has been the goal of all research towards Radio Frequency Identification. Unfortunately, in reality, several problems such as missed readings hinder RFID from being as globally utilised as it should be. In this investigation, we present a /emph{Deferred Bayesian Network} solution that has obtained its network configuration through means of /emph{Genetic Algorithm} training. From this research, we have found the optimal amount of chromosomes required to ensure the highest possible result. Additionally, we have compared our methodology with that of a static bayesian network from previous research and have discovered that our trained algorithm yields a cleaner data set result.
View less >
View more >The alluring promise of a system that would effortlessly identify a mass of objects has been the goal of all research towards Radio Frequency Identification. Unfortunately, in reality, several problems such as missed readings hinder RFID from being as globally utilised as it should be. In this investigation, we present a /emph{Deferred Bayesian Network} solution that has obtained its network configuration through means of /emph{Genetic Algorithm} training. From this research, we have found the optimal amount of chromosomes required to ensure the highest possible result. Additionally, we have compared our methodology with that of a static bayesian network from previous research and have discovered that our trained algorithm yields a cleaner data set result.
View less >
Conference Title
Proceedings of the First Malaysian Joint Conference on Artificial Intelligence (MJCAI 2009)
Publisher URI
Subject
Database Management