codyclonts / zillow_clustering_project

Zillow Clustering Project

Project Planning

• Use information from the zillow data base to acquire, prepare, explore, and make models to predict the logerror of single family residential properties.

Project Goals

Answer the question : What are some of the reasons for log error and how can we reduce it?

• Ask questions during the exploration phase to better understand what could be factoring into the target variable of logerror
  • answer those questions with statistical tests and visualizations
• These answers will be used to hopefully reduce log error in the future
• Construct machine learning models for predicting the logerror of the properties
  • run most effective machine learning model against test
• ensure everything is documented and annotated
• Give 5 mminute presentation to Zillow Data Science Team

Business Goals

• Find key drivers for logerror for single family residential properties
• Deliver a report to the Zillow Data Science Team with adequate documentation and comments
• Construct a machine learning model that can be used to predict the logerror for single family residential properties
• Try to beat the current machine learning model to reduce log error in the future
• This will give more accurate property values to consumers when they are attempting to buy a home

Audience

• Zillow Data Science Team

Deliverables

• jupyter notebook containing the final report
• python modules that can be used to reproduce the work
• scratch notebooks that can be refered to for my work
• live presentation of final notebook
• readme file explaining project

Data dictionary

• factors that were brought from the zillow data set

Project plan

• Acquire the data from the codeup db. Create an acquire.py file
• Clean and prepare the data for the exploration phase. Create a prepare.py file to recreate the work
• Explore the data and ask questions to clarify what is actually happening.
  • ensure to properly annotate, comment, and use markdowns
  • write out each null and alternative hypothesis
  • visualize the data
  • run statistical test on the data
• create at least 3 different machine learning models
• choose the model that performs the best
  • evaluate on test
• Deliver final presentation to Zillow Data Science Team

Initial Hypotheses:

• Logerror has a relationship with the location of the property
• Logerror is strongly related to the time of year that the property was sold.
• Log error is related to the value of the property.
• Log error and the age of the property are related.

Executive Summary

• A few key factors have a relationship with log error, such as:
  • The location of the property
  • When the property was purchased
  • The age of the property
• The value of a property doesn't seem to be a key factor of log error
• It seems that the more properties that are sold during a time frame the lower the log error.

Reproduce the project

• In order to reproduce the project, you will need
  • env.py file that will enable you access to the CodeUp database
  • credentials for codeup database
  • All other files contained in this repository
  • clone this repository to your local machine
  • Understanding of what log error is (can be found in data dictionary)