UniNest: Housing Price Estimator

Medium Artcile:

Data geared project designed to project housing market prices in University cities

Web Scraping

Created a web scraper from scratch using Python and HTML-Requests to scrape the following from over 1200 listings across 6 cities:

• Location
• Title
• Price
• Style
• Bedrooms
• Bathrooms
• Size
• Air Conditioning
• Den
• Bachelor/Studio

Data Cleaning

Cleaned the raw data scraped from Kijiji to use for EDA, graph plotting, and model building. The cleaning included:

• Removing "Wanted" listings
• Streamlined bedrooms, bathroom and price to include integers only
• Parsed through bedrooms to create binary columns for dens
• Removed "Please Contact" listings

Exploratory Data Analysis

Analyzed the clean data and making bar, scatterplot and heatmap graphs using Tableau & Jupyter Notebook [Seaborn, Matplotlib]

The following graphs analyze relationships between the various data columns including Size, Price, Bathrooms, Bedrooms, Location and Den status:

View the Dashboard: https://public.tableau.com/views/HousingPriceBoard/Dashboard1?:language=en-US&publish=yes&:display_count=n&:origin=viz_share_link

Machine Learning Model

Trained a multiple linear regression model with sklearn (accuracy ~70-80%) to predict housing data based off the information I scraped and cleaned. Deployed the model using a Flask API.

Resources & Packages Used

Python Version: 3.11.4

Packages: pandas, numpy, matplotlib, seaborn, html-requests

EDA video: https://www.youtube.com/watch?v=QWgg4w1SpJ8&list=PL2zq7klxX5ASFejJj80ob9ZAnBHdz5O1t&index=4&t=2011s&ab_channel=KenJee