Analysis of patient demographics and test results using data mining methods and thyroid cancer examination

The aim of this study is to detect thyroid cancer and cancer recurrence using 383 datasets containing 16 parameters. These variables are: age, gender, smoking, smoking history, whether got radiotherapy or not, thyroid function status, physical examination, adenopathy, pathology, focus, risk type, T, N, M stages depending on risk type, cancer level and recurrence status. In this study, Decision Stump, Hoeffding Tree, J48, LMT, Random Forest1, Random Forest2, REP Tree trees datasets and Naive Bayes, Logistic Function, Multilayer Perception Function, Simple Logistic Function1, Simple Logistic Function2, IBK K 3 functions were run with WEKA program. According to the results, it is concluded that Random Forest trees are better than other classifiers and studies in the literature.