From the course: Certified Analytics Professional (CAP) Cert Prep

Unlock this course with a free trial

Join today to access over 25,600 courses taught by industry experts.

Running and evaluating models

Running and evaluating models

From the course: Certified Analytics Professional (CAP) Cert Prep

Start my 1-month free trial Buy for my team

Running and evaluating models

“

- [Instructor] From the previous lesson, we learned that data analytics is all about making tough choices just like anything in our lives. However, we data scientists have the luxury of trying out these different models, and techniques before we pick the best one for a given problem. Because there's so many variables that contribute to a model structure, it's extremely difficult to know which combination works best from the beginning. This is why we need to try different models with various settings many times until we arrive at the best solution. Let's revisit our fraud detection example. The goal here is to divide the population of credit card transactions into two groups, that is legitimate and fraudulent. More academic way of saying the same thing is classification with a binary target, which sounds impressive, but may not mean anything to lay people. Decision trees, logistic regression, and neural…

Contents