@Article{, title={Discriminant Analysis of Bronchitis Cancer Data}, author={Nazera Khalil Dakhil نزيرة خليل داخل and Yahya Mahdi Al-mayali يحيى مهدي الميالي and Gahssan Dahair Al-Thabhawee}, journal={Journal of Kufa for Mathematics and Computer مجلة الكوفة للرياضيات والحاسوب}, volume={1}, number={6}, pages={97-106}, year={2012}, abstract={The aim of this research is to predict membership intwo mutually exclusive groups of bronchitis cancerpatients, and allocating new patients usingDiscriminant Analysis.Discrimination is a multivariate techniqueconcerned with separating distinct sets of objects(or observations) and with allocating new objects(observation) to previously defined groups. Theresults showed 90% , and 98% of dead and alivepatients were classified correctly. Only 2% and10% of dead and alive patients were misclassifiedIntroductionDiscriminant analysis was first introduced by R.A.Fisher in 1936 [6]. It is rather exploratory in nature.As a separative procedure, it is often used on a onetimebasis in order to investigate observeddifferences when fundamental relationships are notwell understood. Classification procedures are lessexploratory in the sense that they lead to welldefinedrule, which can be used for assigning newobjects.Previous studies showed the used ofdiscriminant functions to classify three types oflesions in three groups: The normal, the benign,and the malignant. It was observed that thecorrectly classified carcinoma is only 42% andfor normal are 100%. [7].The main contribution of this paper is to proposea simple Fisher-type discriminant method ongene selection in microarray data.[5]Other study compared the performance ofdifferent discrimination methods for theclassification of tumors based on gene expressiondata.[9]

} }