From the course: Advanced Predictive Modeling: Mastering Ensembles and Metamodeling
Unlock the full course today
Join today to access over 22,600 courses taught by industry experts or purchase this course individually.
Voting for classification - SPSS Tutorial
From the course: Advanced Predictive Modeling: Mastering Ensembles and Metamodeling
Voting for classification
- [Instructor] Okay, you may recognize this as the famous Titanic dataset. We're going to use this dataset to look at three straightforward ways of combining your models through voting. Okay, so what we're looking at is we have a CHAID model and it's confidence, a CRT model and it's confidence, and a C5 model and it's confidence. So what's interesting is that sometimes the models don't agree, and naturally that's an important part of why we're doing the ensemble. So, if we take a look at this gentleman, Mr. Moran, we can see that the CHAID model predicts that he's going to die, the CRT model that he's going to survive, and the C5 model that he's going to die. How do we resolve this? Well let's take a look at three different approaches that we can use. We can use simple voting, and two out of three voted for zero, so our simple vote will also be zero. Notice the confidence of our simple vote, it's 2/3. Why? 'Cause 2/3 of the model said zero, and the other one did not, so that's where…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.