From Learning and Evolution to Data Science


  • Home

  • Categories

  • About

  • Archives

  • Tags

  • Search
close

Exploring House Price Estimation And Imagining A Better Real Estate Website

Posted on 2015-10-23   |   In Data Science   |  

Summary

  • This personal project was inspired by my own experience in housing market. I was wondering what I can achieve with public data and Data Science methods.
  • Built upon extensive research of domain knowledge, my model is very simple yet powerful. It used the data with most information, followed the trends in space and time.
  • The prediction accuracy is comparable to Zillow Zestimate, although I don’t really think this accuracy number matters too much. The variation is still too large which is in part due to the nature of house sale/bid process.
  • Instead of trying to improve prediction accuracy numbers, I think it is more practical to list more reasonable comparable sales to user.
Read more »

Simple Python performance timing by checkpoints

Posted on 2015-10-20   |   In Python   |  

Summary

  • This is a simple python script that can measure python program running time in fine levels.
  • It’s simpler than a full profiler, easier to use than other currently available similar scripts.
Read more »
12
dracodoc

dracodoc

This is about my exploration of Data Science, came from research background of AI, artificial learning and evolution.

12 posts
5 categories
21 tags
RSS
Links
  • R-bloggers
© 2015 - 2019 dracodoc
Powered by Hexo
Theme - NexT.Mist