The present work proposes a vibration signal based intelligent bearing fault diagnosis using various prediction models. It includes different feature selection algorithms ReliefF, Information gain, Gini index and random forest algorithm, subsequently classifiers such as JRip,J48,Reduce error pruning, Logistic model tree (LMT), decision table and RIDOR were used to predict the bearing conditions .The experiment were conducted for four cases such as Normal Bearing, Inner race fault, Outer race fault and Ball fault, at constant speed and load conditions and the vibration data is obtained . The aim of the paper to identify an appropriate model for that maintains high accuracy with adequate computational time. It was observed that RIDOR possess higher accuracy than other classifiers and LMT given optimum computational time with decent accuracy prediction. Other parameters related to classification process were also discussed. © August 2016 IJENS.