Bangaluru house price prediction_model

Overview

• Predicting house prices is a crucial task in real estate, helping buyers and sellers make informed decisions. This project leverages machine learning techniques to predict house prices in Bengaluru, India, based on various features.

Objective

The goal of this project is to develop a predictive model that estimates house prices using historical data, considering factors like location, size, number of bedrooms, and amenities.

• Performing data exploration to understand trends.
• Creating relevant features such as price per square foot.
• Removing outliers to improve prediction accuracy
• Optimization model perforamce using Grid Dearch CV
• Using different algorithms to comapre their accuracy respectively

Data Source

• https://github.com/TheMLengineer07/Bangaluru-house-price-prediction-model/blob/main/bengaluru_house_prices.csv

Data Preprocessing

The dataset undersoes preprocessing steps,including

• Feature Addition: Creating Price per square foot as an informative feature
• Oulier Detection & Removal: detect outlier and remove them which creat noise in the model
• Categorical Feature Engineering: Encoding location based features for better predictions

Methodology

Exploratory Data Analysis(EDA)

• Understanding the distribution of price, square footage, and bedroom counts.
• Visualizing price trends across locations.
• Identifying extreme values that act as outliers.

Feature Engineering

• Creating price per square foot for improved accuracy.
• Handling categorical variables using one-hot encoding.

Model Selection

• we use different algorithms like-Linear Regression,Lasso and Decision Tree to compare their accuracy
• Creating price per square foot for improved accuracy.
• Handling categorical variables using one-hot encoding.

Implementation

Tools and Technology

• Programming Language:Python
• Libraries Used:Pandas,numpy,sklearn.matplotlib,Searborn

#Results and Analysis

Model Performance

• Accuracy: Evaluated using R-squared score and find Linear Regression model with 84% has the maximun accuracy.
• Error Metrics: Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE).
• Feature Importance: Understanding how different factors affect price.

Summary

• The project successfully predicts house prices using Linear Regression,Decision Tree Regressor and Lasso
• Removing outliers and feature engineering improves accuracy.
• Hyperparameter tuning using Grid Search CV optimizes model performance.