Application and comparison of different classification methods basedon symptom analysis with traditional classification technique for breastcancer diagnosis

Novel approach for classification technique such as Artificial Neural Network (ANN), Linear DiscriminantAnalysis (LDA) and Random Forest (RF) using factor or dichotomic variables has been introduced. Thisstudy searches for the highly informative finitely linear combinations (symptoms) of variables in the finitefield on the based of the Fisher’s exact test and accurately predict the target class for each case in the data.There are several super symptoms have comparable ???? − ????????????????????????. In this case, it becomes possible to chooseas a nominative representative the factor which is more accessible for interpretation. The super symptommeans a linear combination of various multiplications of k dichotomous variables over a field ofcharacteristic 2 without repeating. In algebra, such functions are called Zhegalkin polynomials or algebraicnormal forms.This process essentially yields the new variable of the identical nature or factor. The purpose of this studyis to suggest a classifiers in accordance with symptoms analysis, and accurately predict the class in thedataset by using different algorithms. The proposed method for super symptoms with the most famousclassification methods was compared to traditional classification methods. The performance of theclassifiers has investigated based on breast cancer data set for training algorithm. Moreover, these threedifferent algorithms have been studied very well based on symptom analysis and thus we do focus on thefact that the best results are from which algorithm.