Mammography is a low dose x-ray procedure for the visualization of internal structure of breast. It detects about 80-90% of the breast cancers without any note of symptoms. A framework for classifying mammograms as tumor and no tumor is presented in this paper. Symlet wavelet and Singular Value Decomposition (SVD) are used for feature extraction and reduction respectively. Boosting algorithm is applied to predictive data mining to generate a sequence of classifiers. A hybrid learning Artificial Bee- AdaBoost (AB-AB algorithm) is proposed by combining concept of Artificial Bee Colony (ABC) algorithm and AdaBoost algorithm. The proposed hybrid algorithm boosts the classification ability of Support Vector Machine (SVM).MIAS dataset is used for evaluating the proposed method. Experimental results are conducted for AdaBoost and proposed optimization technique. © 2015 Elsevier B.V., All rights reserved.