A
Mini Project On
“House price prediction”
Submitted
By
Student Name: Shivani Sunil Matsagar Seat No: B400930132
Student Name: Aniket Bhanudas Jambukar Seat No: B400930120
Student Name: Rushikesh Vijay Bhor Seat No: B400930107
Student Name: Ajay Dinkar Patil Seat No: B400930139
In partial fulfillment of
Bachelor of Engineering
[BE Computer Engineering]
[2025-26]
Department of Computer Engineering
Loknete Gopinathji Munde Institute of Engineering
Education and Research
Nashik, 422002
Loknete Gopinathji Munde Institute Of Engineering
Education and Research
Department of Computer Engineering
Nashik, 422002
CERTIFICATE
This is certify that the Mini Project entitled “House price prediction”, submitted by Rushikesh
Vijay Bhor is a record of bonafide work carried out by him, in the partial fulfilment of the
requirement for the award of Degree of Bachelor of Engineering (Computer Engineering) at
Loknete Gopinathji Munde Institute of Engineering Education and Research, Nashik under the
Savitribai Phule Pune [Link] work is done during year 2025-26.
Prof. [Link] [Link]
(Subject In-Charge) (HOD of Computer Department)
Date:
Place:
CONTENTS
1. Introduction
1.1 Motivation
1.2 Objective/Purpose
1.3 Scope of Project
2. Theoretical Framework and Paradigm Selection
2.1 Transformer Task Selection: Classification vs. Generative NLP
3. Software Requirements
4. Hardware Requirements
5. Implementation Details and Architectural Flow
5.1 Phase 1: Data Preprocessing
5.2 Phase 2: Feature Selection
5.3 Phase 3: Model Traning
5.4 Phase 4: Model Prediction
6. Evaluation Protocol and Result Analysis
6.1 Core Evaluation Metrics
7. Conclusion and Future Outlook
1. Introduction
House price prediction is an important application of machine learning in the real estate sector,
where the goal is to estimate the value of a property based on various features such as area, number
of bedrooms, bathrooms, number of stories, and parking facilities. With the increasing demand for
accurate and quick property valuation, traditional manual methods are becoming less efficient and
time-consuming. By using machine learning techniques like linear regression, it is possible to
analyze historical housing data and identify patterns that influence property prices. This project
focuses on building a predictive model that learns from existing data and provides accurate price
estimates for new house conditions, helping buyers, sellers, and real estate agents make informed
decisions.
1.1 Motivation
The motivation behind the house price prediction project arises from the growing need for accurate,
fast, and data-driven property valuation in the real estate market. Traditional methods of estimating
house prices often rely on human judgment, which can be subjective, inconsistent, and time-
consuming. With the availability of large amounts of housing data, machine learning provides an
efficient way to analyze patterns and relationships between property features such as area, number
of rooms, and amenities. This project aims to leverage these techniques to build a reliable
prediction system that can assist buyers, sellers, and real estate professionals in making better
decisions, reducing uncertainty, and improving transparency in the housing market.
1.2 Objective/Purpose
The main objective of this project is to build a machine learning model that can accurately predict
house prices based on various input features such as area, number of bedrooms, bathrooms, stories,
and parking facilities. The model is trained using historical housing data, allowing it to learn
patterns and relationships between these features and the corresponding prices. After training, the
model’s performance is evaluated using standard metrics such as Mean Absolute Error (MAE),
Mean Squared Error (MSE), and R² score to measure its accuracy and reliability. This ensures that
the model provides meaningful and dependable predictions for real-world applications.
1.3 Scope of Project
The scope of this project is focused on developing a machine learning-based system for predicting
house prices using a structured dataset containing features such as area, number of bedrooms,
bathrooms, stories, and parking spaces. The project involves data preprocessing, model training,
prediction, and evaluation using appropriate performance metrics. It is limited to using simple
regression techniques like Linear Regression and does not include advanced real-world factors
such as location mapping, economic conditions, or real-time data updates. Additionally, the project
may include a basic user interface for input and prediction, but it does not cover full-scale
deployment or integration into commercial real estate platforms.
2. Theoretical Framework and Paradigm Selection
The theoretical framework of this project is based on the principles of supervised machine learning,
specifically regression analysis, where the goal is to predict a continuous output value in this case,
house price based on multiple input features. The model used in this project is Linear Regression,
which assumes a linear relationship between independent variables such as area, number of
bedrooms, bathrooms, stories, and parking, and the dependent variable, price. The algorithm works
by finding the best-fit line that minimizes the error between actual and predicted values using
techniques like least squares. The dataset is divided into training and testing sets to ensure that the
model can generalize well to unseen data. Performance is evaluated using metrics such as Mean
Absolute Error (MAE), Mean Squared Error (MSE), and R² score, which provide insights into the
accuracy and reliability of the model. This framework ensures a systematic approach to building,
training, and evaluating the predictive model.
3. Software Requirements
• Python 3.x
• Jupyter Notebook
• Libraries:
o Pandas
o NumPy
o Matplotlib
o Scikit-learn
4. Hardware Requirements
• Processor: i3 / i5 or higher
• RAM: 4GB minimum (8GB recommended)
• Storage: 1GB free space
5. Implementation Details and Architectural Workflow
The practical realization of this project involves a meticulously structured, multi-phase pipeline.
The workflow transitions raw visual data into structured numerical features, serializes those
features into linguistic sequences, and finally fine-tunes a pretrained transformer to act as a robust
classifier. The following subsections exhaustively detail the programmatic logic and architectural
configurations required at each milestone, explicitly highlighting where code implementation
screenshots should be documented within the final deliverable.
5.1 Phase 1: Data Preprocessing
Steps:
• Load dataset
• Handle missing values
• Convert data into usable format
df = pd.read_csv("[Link]")
[Link]([Link](),
inplace=True)
5.2 Phase 2: Feature Selection
Input features:
• Area
• Bedrooms
• Bathrooms
• Stories
• Parking
Target:
• Price
5.3 phase 3: Model traning
from sklearn.linear_model import LinearRegression
model = LinearRegression()
[Link](X_train, y_train)
5.4 phase 4: Prediction
price = [Link]([[2000, 3, 2, 2, 1]])
6. Evaluation Protocol and Results Analysis
Metrics Used:
• MAE (Mean Absolute Error)
• MSE (Mean Squared Error)
• R² Score
from [Link] import r2_score
print(r2_score(y_test, y_pred))
6.1 Core Evaluation Metrics
The core evaluation metrics used in this project are essential for measuring the
performance and accuracy of the house price prediction model. Since this is a regression
problem, metrics such as Mean Absolute Error (MAE), Mean Squared Error (MSE), and
R² Score are used. Mean Absolute Error (MAE) calculates the average absolute difference
between the actual and predicted house prices, providing a clear idea of how much error
the model makes on average. Mean Squared Error (MSE) measures the average of the
squared differences between actual and predicted values, giving more importance to
larger errors and helping identify significant deviations. The R² Score, also known as the
coefficient of determination, indicates how well the model explains the variability of the
target variable, with values closer to 1 representing better performance. Together, these
metrics provide a comprehensive understanding of the model’s effectiveness and
reliability in predicting house prices.
Fig 6.1: Prediction graph
7. Conclusion and Future Outlook
The model successfully predicts house prices by learning the relationship between input features
such as area, number of rooms, and other factors, and the target variable, which is the price. Linear
Regression proves to be effective for this project, especially since the dataset is simple and shows
a relatively linear relationship between features and output. It provides reliable and interpretable
results, making it suitable for a basic implementation. However, the model can be further improved
by using advanced machine learning algorithms such as Random Forest, Decision Trees, or
Gradient Boosting, which can capture more complex patterns in the data and potentially increase
prediction accuracy.