Classifying clinically actionable genetic mutations using KNN and SVM

Cancer is one of the major causes death in humans. Early diagnosis genetic mutations that cause cancer tumor growth leads to personalized medicine decease and can save life majority patients. With this aim, Kaggle has conducted a competition classify clinically actionable gene based on clinical evidence some other features related mutations. The dataset contains 3321 training data points be classified into 9 classes. In work, an attempt made these using K-nearest neighbors (KNN) linear support vector machines (SVM) multi class environment. As are categorical, hot encoding as well response coding applied make them suitable classifiers. prediction performance evaluated log loss KNN performed better with value 1.10 compared SVM 1.24.