Cancer Detection Using adaptive Neural Network
Seema Singh, Sunita Saini, Mandeep Singh · 2012
We present a system which detects the cancer stage using clustering techniques. The task is accomplished using Adaptive Resonance Neural Network (ARNN), a special case of unsupervised learning. A vigilance parameter (vp) in ARNN defines the stopping criterion and hence helps in manipulating the accuracy of the trained network. To demonstrate the usefulness of ARNN, we used Wisconsin breast cancer database. The database available in the UCI database repository contains 699 cases out of which we used 600 cases to train the network. In this dataset there are 375 benign cases and 225 as malignant cases. We see that at vp=0.2 the network has Recall (i.e. true negative rate) is 75% and average Accuracy=82.64% and Precision is 79%.