Data Analytics

As part of my General Assembly project submission, I took a Coursera dataset to evaluate possible courses to take next for self-paced learning.

Objectives

Coursera believes that they need to increase their hands-on training centers. They also believe that they need to invest in teacher training programs to improve some of their courses. “We believe that higher ratings for our instructors will result in more enrollments”

Hypothesis 

• Hands-on courses do better than flexible courses

• The higher the average rating, the more the enrollment in that course 

• Poorly performing instructors are the reason for poor ratings 

• Data Cleaning

• Removed 1,164/33,530 “0” rating rows from the data set

• Removed 16/8,557 courses with no title 

• Categorized duration 

• Categorized Ratings

• Calculated weighted average of ratings based on the number of reviews 

• I only looked at the top 10 instructors and their performances