Udacity: Classification Models Review

Classification Models is a short self-paced introduction to data prediction tasks with non-numeric targets offered by Udacity in collaboration with Altreryx as a part of its business analyst nanodegree program. The course is free and does not require coding knowledge since you'll use the Alteryx software used throughout the nanodegree program for any data analysis tasks. The course home page suggests that the course takes approximately 2 weeks to complete, but won't take a focused student more than a few hours to get through it.


The course is split into 3 sections: an overview of classification problems, a section on binary classification and a section on non-binary classification. Each section has roughly 25 lesson segments that consist of lecture and demonstration videos accompanied by text highlight key points and occasional quiz questions. The video segments are very short, averaging around 2 minutes or less. The videos have good production value and explain concepts well, albeit in fairly shallow terms given their brevity.  Still, the course manages to touch on some of the most popular classification methods like logistic regression, decision trees, random forests and boosting and demonstrate how to use them in Alteryx.


Udacity's Classification Models is an accessible introduction to predictive modeling for classification problems that feels more like a quality tutorial than a full course. You could easily complete the course in one sitting. If you aren't taking other courses in the nanodegree program, you could take this course to learn the general concepts and skip the Alteryx demonstration segments.


I give Classification Models 3.5 out of 5 stars: Good.