Alireza khosravi - Babol university ofTechnologyFaculty of Electrical and ComputerEngineeringBabol, Iran
Jalil Addeh - Babol university ofTechnologyFaculty of Electrical and ComputerEngineeringBabol, Iran
Javad Ganjipour - Babol university ofTechnologyFaculty of Electrical and ComputerEngineeringBabol, Iran

This paper presents an accurate hybrid system forrecognizing breast cancer tumours which includes three mainmodules: feature extraction module, training module andclassifier module. In feature extraction module, fuzzy feature hasused as an effective classifier input. In training module, ahybrid bees algorithm (BA) - back-propagation (BP) algorithm isproposed to train the classifier. This module enjoys theadvantages of global search of BA and local search of BPalgorithm. In classifier module, multi-layer perceptron (MLP)neural network is used. The proposed system is tested onWisconsin breast cancer (WBC) database and the simulationresults show that the recommended system has high recognitionaccuracy in comparison with other methods.

Back propagation, Bees algorithm, Breastcancer, Neural network; Fuzzy feature