A novel approach on grinding wheel wear prediction to detect a worn out wheel is proposed in this paper, using PCA-DA, PLS-DA and Naïve Byes data mining techniques from acoustic emission (AE) signals processed by discrete wavelet transform. The statistical features were extracted from the wavelet co efficient carried out for each wavelet decomposition level.The proposed approach was validated with AE signal data collected in Aluminium oxide 99A(38A) grinding wheel which is used in many of the common grinding operations in general practice under different grinding conditions. Validation results of the proposed data mining techniques for different machining conditions with respect to classification accuracy were discussed. © Research India Publications.