Please enable JavaScript.

Coggle requires JavaScript to display documents.

MACHINE LEARNING (SUPERVISED (REGRESSION (Accuracy (R-squared: Normalizado…

- - - - R-squared: Normalizado
        
        summary(model)$r.squared
      - RMSE: No está normalizado. No vale
  - - - conf<-table(test_labels, pred)
      - accuracy <- sum(diag(conf))/sum(conf)
      - normalize :: (knn_train$Pclass - min_class) / (max_class - min_class)
      - pred<- knn(train = knn_train, test = knn_test, train_labels, k = n)
    - - tree<- rpart(spam ~ ., train, method = "class", parms = list(split = "information"), control = rpart.control(cp=0.00001))
      - pred<- predict(tree, test, type="class")
      - conf<-table(test$spam, pred)
      - accuracy <- sum(diag(conf))/sum(conf)
      - fancyRpartPlot(tree)
      - pruned<-prune(tree, cp = 0.01)
    - - prediction(probs_k,test$spam)
      - performance(pred_k,"tpr","fpr") / (pred_k,"auc")
      - draw_roc_lines(perf_t,perf_k)