Cancer.R file: Prediction of breast cancers diagnostics in Wisconsin. I compare the prediction performance of three methods: Random Forests, SVM and Logistic Regression. I present a way to get the truly best model by looking at accuracy distribution and IQR.

Data.csv file: Contains the data, originally found here https://www.kaggle.com/uciml/breast-cancer-wisconsin-data. It contains 569 observations and 32 independent variables. The dependent variable, diagnosis, is binary: malign cancer=1 or benign cancer=0.

Logistic Regression performs best, both in average accuracy and in the distribution of the accuracy.

Article presenting the analysis available here: https://towardsdatascience.com/a-simple-way-to-pick-the-right-model-d362272b453d