Project 3 Learning and Improvements
The project of involved leveraging machine learning models to predict the number of shares for a particular channel based on the features provided. I worked with Brant Armstrong on this project who was a great partner and helpful. He really showed empathy when I was sick and helped me through the project, really appreciate his partnership as part of this project.
What would you do differently?
If given the opportunity to work on this project again, I would do the following different:
1) I would first of all spend some time reading the theories of the models a little more in depth. That would allow me to really thing about which model would work best with the data.
2) Second, I would spend some more time understanding the data by leveraging exploratory data analysis.
3) Third, spend some more time with variable selection to see how can the data be used best to extract the most out of it.
What was the most difficult part for you?
The most difficult part for me in the project was the automation of the project. I tried implementing the render function, but I failed to incorporate a small feature which would allow each .md file to be pickup the respective data, this caused each of our files to use the same data from the lifestyle channel.
What are the big takeaways from the project?
The biggest takeaway from the project was the automation part again, and how each channel would perform differently for each model. Firstly, the automation was very interesting to see and how everything came together once the render function finally worked. Secondly, it was interesting to see how each model provides a different result and how the data reacts to each of the models. But all in all, a good hands on with automating and working with different models.
Additionally, I just want to thank Dr. Post and Brant, my group member, for there understanding and extreme helpful nature. I was very worried that I would not be able to submit the project on time. Thank you so much to both of them for being such great humans.
Finally, like always, oepn to all comments and suggestions towards the work presented.
Repo: https://github.com/Branticus71/Project-3-Group-A
Project Report: https://branticus71.github.io/Project-3-Group-A/LifestyleAnalysis.html
Sincerely,
Naman Goel