AnalyticsVidhya has 2 great posts from 2015 about how to quickly build a basic predictive model in 10 minutes. One post is for R and the other for Python. As I am more focused on helping introduce people to data science, let’s say building a model in 30 minutes instead of 10. The idea behind creating a fast, basic model is one I tend to agree with when you are first exploring a data set and what it can tell you about your question for the data. That idea is “quickly establish a benchmark, then refine your model”. As you begin to get more used to building models, you’ll find that these quick first builds do fairly well. They also give you great insight into what variables are important, what is not, and what still needs to be refined and cleaned up. All of this is great practice to help you once you are out doing this in a non-learning environment. Run through these with a few data sets you have available and see what you find!
R link to article. Python link to the article.
(If the links don’t work the urls are:
R – https://www.analyticsvidhya.com/blog/2015/09/perfect-build-predictive-model-10-minutes/
Python – https://www.analyticsvidhya.com/blog/2015/09/build-predictive-model-10-minutes-python/)
Image taken from Credera post by Elizabeth Jones and Cameron Randall