Would love to hear feedback from my peers or those who are more advanced in this topic

This is the link to my code: https://github.com/LiamJ95/U.S.-Medical-Portfolio-Project-/blob/main/us-medical-insurance-costs.ipynb

This is my first time sharing anything, so I apologize if I did not share it correctly.
I am around 27% complete of the Data Scientist: Machine Learning career path.

Nice. Some things I’d be interested in at the end:

Variance (and maybe skew). Reasoning: Sure, smokers may cost more on average. But how consistent is the cost for a non-smoker. From the point of view of cost-savings this could be a consideration when there are other cost-saving opportunities. For skew, I want to know, when it varies, does it vary on the higher end on lower end, or right in the middle? Are there regional trends with these (var and skew)?


Congrats on completing the project. I forget if at this point in the course you’ve been introduced to Matplotlib or Seaborn?

A few things:

  • don’t put your conclusions at the top of the notebook, put them in a nice summary at the end of your analysis. The intro should just have a cite as to the data source and what you hope to look at in the data (which you partially have already).

  • I like that you broke down the regional costs into mean, median, min, & max.

  • there are outliers in the data so that will pull the mean. Better to look at the median in regards to charges.

  • slight detail: in re: “Smokers in the Southeastern Region make up 25.0 % of all the customers in this Region which adds to why it may cost more to insure customers here”
    The data set isn’t the costs of insuring people; it’s the amount of money individuals pay out of pocket (in the U.S.) for medical care (I’m assuming before their insurance actually kicks in).

  • in the section, “Relationships between the data”, you haven’t done any hypothesis testing between the variables, so, you can’t see any possible correlations (yet) in the data. You could say something like, the data shows ____ and ____", etc.

Good work. This is a project that you will come back to in the course.

1 Like

I want to say first and foremost, thank you for the valuable insight. I’ll be honest, I thought I would get more info about the code and completely forgot about the most important part, relaying the information back in an easy-to-digest format and just being correct in what I am trying to say. So again, thank you very much.

1 Like