2
I would like to know how to evaluate which model is better. For example: I am learning machine Learning using the dataset the Titanic, and I used Tree decisions, Logistic regression and Random Forest, but how can I assess which will have the highest percentage of hits??
I read a material on machine evaluation metrics Learning, but I do not understand very well and even less how you use.
preferably using R
Pedro, this site gets better questions about more objective programming. Your question is very open and a good answer would require a very large text. Try to edit your question by explaining exactly what you don’t understand: "what are they for?" "what’s the idea behind it?" This will make it easier to get an answer. In principle the very "percentage of hits" you quoted is an evaluation metric. Also try to post the code you are using to adjust these templates you mentioned.
– Daniel Falbel
I would like to know how to use the evaluation metrics
– Pedro Teixeira
Somewhere where I can study the subject, it would be something very good
– Pedro Teixeira
Have you tried it here: https://www.kaggle.com/amberthomas/titanic/predicting-survival-on-the-titanic
– Daniel Falbel
In this link has a plethora of R scripts analyzing this database.
– Daniel Falbel
Hi. Welcome to SOPT. As Danielfalbel already explained to you, your question is not very good for the format of this site (which is not a forum - if not yet, do the [tour] and read [Ask]). You say you "read material about metrics, but didn’t understand". Don’t understand what? Try to explain your doubt!
– Luiz Vieira
@Danielfalbel, thanks for the link, already opened me a new horizon
– Pedro Teixeira