A venture capital fund wants to obtain a criterion to invest in fifty startups. Based on a dataset containing: their expenditure on R & D, administration and marketing, location (state), and profit, we will find which features are significant to these startups' profits.
You will also find an R file with the equivalent code, which is optional. This R code contains three different ways to implement Backwards Elimination. The first one consists of running the different linear models so the user can compare them. The other two ones are:
-
1. The use of the R function step().
-
2. It is a function where an algorithm for Backward Elimination has been implemented. A significance level - SL- of 5% has been pre-set (you can copy it for your projects and experiment with other SL).
I have also added the R code with a Google Colab R runtime. So, if you are interested in knowing how to do it, please visit this site by Fidocia Wima Adityawarman, who explains how to easily change the runtime for using R in Colab:
Note: If the Colab notebook is not being displayed, please copy the URL and paste it on nbviewer so you can see the code.
This was not part of the main task for this project, but I have added a file in Python so you can see the descriptive statistics, histograms, and boxplots.
Please note that this code is intended for educational and non-commercial use only.
Contributions to this repository are welcome. If you find a bug or have suggestions for improvement, please open an issue or submit a pull request.
This project was created by Santiago Moreno Velasquez as part of an Udemy Guided Project.