Regression-Model-Project

This project uses the zillow dataset on MySql through Codeup's servers and produces a regression model to predict tax value.

Goal

• The purpose of this model is to predict tax values of single family properties using features of the zillow dataset from 2017.
• My goal is to find specific features that are correlated with the tax value and build a model with lowest RMSE to predict property tax values on new data.

Initial hypotheses

• Null Hypothesis:
• Alternative Hypothesis:

Data dictionary

Feature	Description
bedrooms	total bedrooms
bathrooms	total bathrooms
square_ft	square footage
tax_value	tax value
year	year built
county	location

Planning:

Questions to ask about the data set based off of what I want my model to predict:

• Do any features have a correlation with tax value?.

• What features significantly affect tax value?

• Final report should be in .ipynb, Modules should be in .py, Predictions should be in .csv.

• Audience will be lead data scientist.

• Determine correlation between features and tax value.

• Develop my null hypothsisis and alternative hypothesis.

• Explore data using visuals and statistical tests.

• Determine what model to create.

Acquisition:

• Using a function with my credentials stored in a env.py file, I connected to codeup's MySQL server.
• I created a query pertinent to the project.
• Using a function I bult containing the query and pandas.read_sql, I called the function to acquire and read zillow data from MySQL.

Preparation:

• Perform tasks such as handling null values, renaming columns, normalizing text, binning of data, changing data types, mapping column values.
• Open wrangle.py file and look under clean_zillow() function for more details.

Exploration & pre-processing:

• Made visuals and used stats to understand which features had a significant correalation, relationship
• I dropped features that had no significance
• Open explore.py and preprocess.py file for more details.

Modeling:

• I choose a classification random forest model
• I used my train data set to train my model
• I made predictions with my model
• I made multple models and choose the best one

Delivery:

• Deployed my model and a reproducable report
• Made recommendations

Key findings, recommendations, and takeaways

• Monthly contracts amd charges are key features driving customer churn
• I recommend incentivizing yearly contracts by offering promotion deals and a loyalty program

Instructions or an explanation of how someone else can reproduce project and findings

Enviroment setup:

• Install Conda, Python, MySql, VS Code or Jupyter Notebook
• Clone this repo to your local