Cancer disease diagnosis with machine learning methods

The advances and discoveries in Artificial Intelligence in recent years, the concept of Artificial Intelligence has started to be used in almost all fields of life. The number of studies in machine learning, a sub-branch of artificial intelligence, has significantly increased, especially in the field of healthcare. Today, with the increasing number of people dying from cancer, the significance of machine learning studies in this field is growing. Early and accurate cancer diagnosis is crucial for effective treatment and improving survival rates. This study focused on using machine learning techniques for diagnosing and classifying breast cancer using four different datasets. In the first dataset, PCA, LDA, and t-SNE dimensionality reduction techniques and Logistic Regression, Support Vector Machines, Random Forest, Naive Bayes, Gradient Boost, k-nearest Neighbor, Decision Tree, AdaBoost, Extra Gradient Boosting, and Extra Gradient Boost (Random Forest) machine learning methods were used. In the second dataset, Random Sampling and SMOTE resampling techniques were applied and Random Forest, AdaBoost, Gradient Boosting, k-Nearest Neighbor, Decision Tree, Naive Bayes, Logistic Regression, Support Vector Machines and Extra Gradient Boosting machine learning methods were applied. RFE, RFECV, and SelectKBest feature selection methods, resampling techniques, and various machine learning methods were used in the third dataset. In the fourth dataset, GridSearchCV and RandomizedSearchCV, which are hyper parameter optimization methods, were applied, followed by the analysis of various machine learning methods. Different applications were performed on four different breast cancer datasets, and the effects of dimensionality reduction, resampling, feature selection, and hyper parameter optimization on the results of classification methods were investigated. The different techniques and methods used in the study are compared with each other and with other literature studies.