I am new to genetic algorithm.In my project i have taken 140 MRI brain images of which 100 image(50 normal and 50 abnormal) are used for training of classifier and rest 40(20 normal and 20 abnormal) for testing purpose.
Different features are taken: wavelet coefficients, symmetry, energy, entropy, correlation, contrast, area, perimeter, circularity. now i want to obtain a better subset using genetic algorithm.
For that i have consider an initial population first, as string of 1's and 0's(1 for feature included and 0 for feature not included). The fitness function given in the research paper is accuracy+specificity+sensitivity.
Can someone please tell me how i will get these 3 parameters at feature selection stage???