Skip to content

Uploaded Test Slides#12

Open
BenBrandt1 wants to merge 4 commits intoernbilen:mainfrom
BenBrandt1:main
Open

Uploaded Test Slides#12
BenBrandt1 wants to merge 4 commits intoernbilen:mainfrom
BenBrandt1:main

Conversation

@BenBrandt1
Copy link
Copy Markdown

No description provided.

@QuangMinh07112003
Copy link
Copy Markdown

Hi Ben,

Here is my opinion of your idea 1:
Your idea is overally great, and your plan to execute the project is great as well. You have fully stated all parts in the rubric, you've stated all relevant variables that can help predict your dependent variable, which is great. If you still have time, I'd recommend adding some clustering analysis, or building a nice interactive visualization dashboard. That would help your project stand out from others. I assume your data will include locations, so one thing you can do is spatial analysis. One thing to double-check is the terms and services of scraping data on Zillow, not every website is cool with people scraping data off it. Overall, you have a solid plan for idea 1, and I think you're good to go!

@QuangMinh07112003
Copy link
Copy Markdown

Ben,

Here's my feedback on your idea 2:
I think this is another interesting topic that you're trying to make it happen, though I think that this is a bit too much of work for a mini-project. It is still interesting, don't get me wrong, but I presume that you'd need AI, deep learning and some other tools to retrieve the data. Unless you already have those tools and can execute the data retrieval within short amount of time, it's all good, but if not, timing is a thing to consider. II'm not an NBA fan, but I assume they have data of shot locations on SofaScore, they kinda have a heatmap for every player, including shots made, total and missed. Otherwise, I think you've had a clear plan with your model, and you're very good to go at this point!

Good luck with your mini-project!

Created Mini Project Presentation. Other files are parts of the extra libraries added to the file.
@ernbilen
Copy link
Copy Markdown
Owner

One city is no problem. 200 is a little few.. Maybe could expand to 2k or so?
Again Zillow has already done the work you did (which is okay) So your findings are great, but it would be really interesting if your results correlated with theirs and see if yours or theirs have a better fit for house listings. If zestimates are absolutely not possible to scrape, tax values from tax records could work?
It could be nice to plot maps and show each home as a point on the map with general statistics.
You are also missing explanatory variables like, location, school zones etc.
Good ethics discussion!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants