Planning a Property Price Prediction Project

Getting a dang hold of this dang housing market

To put it in brief, this project is the child of idle hands and frustration. I have been looking at the housing market for a while and the only thing I have found out is that I don't know crap about it and nobody around me has any clue how it's actually going to develop nor what's a good property to buy.

So I figured, let's go ask someone or something smarter, after which I looked up some experts in the field and found out what they thought, and they didn't seem to be aligned either.

Therefore I decided that it was time to pick up the task myself, and hand it over to a computer.. or rather.. a machine.

Enough foreshadowing; this project is a machine learning project where me and my friend tries to solve our problem of not understanding the housing market as a whole nor the individual listings.

And like any other diligent data science engineer (I hope this isn't a protected title) my friend and I etched out the challenges that was lying ahead of us:

  1. Choosing a dang good framework
  2. Getting some dang useful data
    1. Find one or more dang good data sources
    2. Collect enough dang data from the sources (We had to scrape our data from sources that removed sold properties, so this project required frequent scrapings - other problems might need but one scraping)
    3. Clean that dang data so that the needle doesn't jump when the machine has to use it
    4. Transform the data into something sensible to the purpose of this project
  3. Understand the dang data
    1. Visualize the data
    2. Use standardized tool to get a better understanding of how our data correlates
    3. raise NotImplementedError
  4. Propose a dang model
    1. raise NotImplementedError
  5. Implement the dang model
    1. Possibly adjust the data
  6. Train the dang model
  7. Predict some dang stuff
  8. Possibly reiterate or pivot the dang thing

Choosing a dang good framework

raise NotImplementedError

Getting some dang data

We couldn't find any public sources that had listing data

  1. We So essentially there's a few elements for this process: !insertgraphics

raise NotImplementedError