Prediction of House Selling

Project Overview

This project aims to predict whether a house will be sold based on various features such as price, location, amenities, and community factors. Using a machine learning pipeline, the project involves data preprocessing, exploratory data analysis (EDA), outlier detection, and model building to achieve reliable predictions.

---

Table of Contents

1. Introduction
2. Dataset Description
3. Key Steps
4. Models Used
5. Results and Insights

---

Introduction

House selling predictions are critical for real estate businesses to understand market trends and improve decision-making. This project uses machine learning techniques to predict sales status (Sold) using a dataset of housing features.

---

Dataset Description

The dataset contains 18 features describing properties and their surroundings:

• Numerical Features: price, room_num, age, n_hos_beds, n_hot_rooms, etc.
• Categorical Features: waterbody, bus_ter, airport.
• Target Variable: Sold - whether the house was sold (1=Yes, 0=No).

---

Key Steps

1. Data Importing and Preprocessing

• Loaded data using pandas.
• Handled missing values and inconsistent data entries.
• Scaled numerical variables and encoded categorical variables.

2. Exploratory Data Analysis (EDA)

• Identified skewness and outliers in features such as n_hot_rooms and rainfall.
• Visualized categorical variables to analyze their impact on the target variable.
• Observations included:
  1. Missing values in n_hos_beds.
  2. Skewness in age and n_hot_rooms.
  3. Uniform values in bus_ter.

3. Outlier Detection and Treatment

• Used statistical methods and visualization tools (e.g., boxplots) for outlier detection.
• Applied transformations and imputation to handle outliers.

4. Model Building

• Built machine learning models to predict Sold using algorithms like:
  • Logistic Regression
  • K Neartest Neighbor

---

Models Used

• Logistic Regression: For baseline binary classification.
• K Neartest Neighbor

---

Results and Insights

• Feature Importance: Identified key factors influencing sales, such as proximity to amenities and property age.
• Model Performance: Achieved high accuracy with ensemble models.
• Actionable Insights: Recommendations for real estate stakeholders to improve property marketing.

---