A Relative Analysis of Multi-Relational Decision Tree Learning Algorithm

This paper provides a comparative delve of the working and execution of MRDT algorithm with that of MRDTL-2, which works on the theory initially propounded by Knobbe et al. This paper also delineates some of the foible of MRDTL viz. calculation speed, exactitude and most importantly handling of missing values. We had used some of the real world data sets from multiparous data mining sweepstakes and accomplished a graphical comparison for the forenamed two approaches. Conclusion from the experiments implies that MRDTL-2 is convincingly more efficacious approach than its forerunners.