🏠🏠 USA House Price Prediction Via Linear Regression 🏷️🏷️

Badge source

This project will follow the Business Analysis (BA) workflow to address house price prediction using linear regression techniques. The business problem is creating a regression model that can accurately predict house prices based on the provided features. Therefore, real estate agents can utilize this model to evaluate the property.

├── Image                       
│
├── Code_USA_House_Price_Prediction.ipynb             <- code
├── LICENSE                                           <- MIT license
├── README.md                                         <- read me
├── housing_price_dataset.csv                         <- dataset
├── USA_House_Price_Prediction_Report                 <- presentation

1. Business Understanding

The goal of this project is to develop a predictive model for housing prices based on various property attributes, including area, number of bedrooms and bathrooms, etc.

2. Data Understanding

The housing dataset was loaded via Colab. The dataset is from Kaggle: https://www.kaggle.com/datasets/muhammadbinimran/housing-price-prediction-data (also please see housing_price_dataset.csv attached). Basic data analysis was performed to identify the shape of data, get column names, find missing values, and generate descriptive statistics. The pair plot demonstrated the relationship between variables. The distribution of the target variable was shown.

• Data Dictionary

Name	Modeling Role	Measurement Level	Description
SquareFeet	input	int	Square Feet of the house
Bedrooms	input	int	Amount of bedrooms
Bathrooms	input	int	Amount of bathrooms
Neighborhood	input	obj	Area neighborhood where the house is
YearBuilt	input	int	Which year it was built
Price	input	boolean	The price of the home

3. Data Preparation

1. Define variables (X and y)
2. Split the data into train and test datasets; 40% of test data
   train: 30000 data
   test: 20000 data
   

4. Modeling

Several machine learning models were included:

• Linear Regression

5. Evaluation

MSE in test data: 2193848415.96 R2 score: 0.57