Pre-owned Car Price And Life Prediction
Using Machine Learning
Team Members :
B644 Surabh Sawant
B645 Raj Sawarkar
B647 Aman Sharma
B680 Simran Yadav
Project Guide :
Prof. Dipak Gaikar
TABLE OF CONTENTS
1. Abstract
2. Introduction
3. Literature Survey
4. Existing System
5. Problem Statement
6. Objective
7. System Architecture
8. Proposed Flow Diagram
9. System Requirements
10.Output
11. Conclusion
12.References
2
ABSTRACT
This project aims to develop a pre-owned car price prediction model using machine learning algorithms such
as Linear Regression, Lasso Regression, Decision Tree, and Random Forest. The model considers parameters
such as model, year of manufacture, fuel type, and kilometers driven. Once the final model is developed, it is
deployed on a Flask website that allows users to input a car's parameters and obtain an estimated price..
The project provides a reliable and convenient way to determine the estimated value of a pre-owned car,
which can aid in making informed buying and selling decisions. This demonstrates the potential applications
of machine learning algorithms in predicting pre-owned car prices in the automotive industry.
INTRODUCTION
This project will determine the price and lifespan of car.
The value of used cars depends on various important parameter that is the age of the car, model, origin,
mileage and horsepower.
Fuel prices are increasing rapidly day by day. Some other factors such as the type of fuel it uses, the interior
style, the braking system, acceleration, the volume of its cylinders, its size, paint color, consumer reviews,
whether it is a sports car, its physical state, whether it is automatic or manual transmission, etc.
As we can see, the price depends on a large number of factors. Unfortunately, the information about these
factors is not available very precisely and the buyer make the decision to purchase the car at a certain price
based on few factors only.
Regression Algorithms gives us output with continuous value and instead of categorized value due to which
it will be predictable to achieve the actual price of car instead of the price range of a car, so they are used.
LITERATURE
SR
NO
AUTHOR PAGE
TITLE
SURVEY
JOURNAL METHODOLOGY FINDING’S
1 Praful Rane, Used Car International There are two primary phases in the This paper compares 3
Deep Pandya, Price Research system: different algorithms
Dhawal Kotak Prediction Journal of for machine learning :
Engineering and 1. Training phase: The system is Linear Regression,
Technology trained by using the data in the data Lasso Regression and
(IRJET),2021 set and fits a model (line/curve) Ridge Regression
based on the algorithm chosen
accordingly.
2. Testing phase: The system is
provided with the inputs and is
tested for its working. The accuracy
is checked
LITERATURE
SR
NO
AUTHOR PAGE
TITLE
SURVEYMETHODOLOGY
JOURNAL FINDING’S
2 Ketan Agrahari, Used Car International System fills in the gaps by The authors of this
Ayush Price Journal of providing buyers and sellers study compared
Chaubey ,Mamoo Prediction Innovative Research with an estimate of the car's Linear Regression to
r Khan, Manas Using in value based on the below Lasso Regression. The
Srivastava Machine Technology(IJIRIT), algorithm available for price data for this study was
Learning 2021. prediction i.e. Linear Regression gathered from Kaggle
and Lasso Regression. and then analysed
using the Python
programming
language.
LITERATURE
SR
NO
AUTHOR PAGE TITLE SURVEYMETHODOLOGY
JOURNAL FINDING’S
3 Pattabiraman Used Car Price International Training and Testing Data The The prediction error rate
Venkatasubbu,Mu Prediction Journal of data is split into training(70% - of all the models was
kkesh Ganesh Using Engineering and 563 records) and testing(30% - well under the accepted
Supervised Advanced 241 records) data sets through 5% of error.
Learning Technology random sampling (seed was set Multiple and Lasso
Techniques (IJEAT),2019. to 2786). Regressions are better at
predicting price than the
Regression Tree.
EXISTING SYSTEM
Existing System includes a process where a seller decides a price randomly and buyer has no idea about
the car and it’s value in the present day scenario. In fact, seller also has no idea about the car’s existing
value or the price he should be selling the car at. To overcome this problem we will develop a model
which will be highly effective.
PROBLEM STATEMENT
The price of a new car in the industry is fixed by the manufacturer with some additional costs incurred by
the Government in the form of taxes. So, customers buying a new car can be assured of the money they
invest to be worthy. But, due to the increased prices of new cars and the financial incapability of the
customers to buy them, Used Car sales are on a global increase. So, this Project will predict the price and
lifespan of used car using various machine learning models. This will enable the customer to make a
decision based on different factors. It will provide the solution to these problems through which we will
get to know which of the factors are significant and tell us how they affect the car's worth in the market.
OBJECTIVE
The research objective of this is to predict used cars prices and its lifespan by scraping data from websites
that sell used cars and analysing the different aspects and factors that lead to the actual used car price
valuation. To enable consumers to know the actual worth of their car or desired car, by simply providing
the program with a set of attributes from the desired car to predict the car price.
To develop an efficient and effective model which predicts the price of a used car according to user’s
inputs. To achieve good accuracy.
To develop a User Interface which is user-friendly and takes input from the user and predicts the price.
SYSTEM
ARCHITECTURE
CAR DATA COLLECTION DATA PRE-PROCESSING TRAIN-TEST SPLIT
DISPLAY THE MACHINE
RESULT LEARNING MODEL
PROPOSED FLOW DIAGRAM
START
COLLECT DATASET
DATA PRE PROCESSING
USING MACHINE LEARNING ALGORITHMS FOR
PREDICTION
Linear Lasso Random
Decision Tree Forest
Regression Regression
SELECT THE BEST MODEL
DISPLAY THE PREDICTED PRICE
SYSTEM REQUIREMENTS
HARWARE REQUIREMENTS
Processor Dual core 2.4 GHz (i5 or i7 series Intel
processor or equivalent AMD)
Memory 4GB
Hard disk Minimum 1TB
SOFTWARE REQUIREMENTS
IDE Jupyter Notebook
Operating system Windows 7 and above
Programming Language Python, HTML, Bootstrap
OUTPUT
1. Linear Regression
16
2. Lasso Regression
17
3. Decision Tree
18
4. Random Forest
19
Website implementation
20
Website implementation
21
COMPARATIVE STUDY OF ALL THE ALGORITHMS
Algorithm Accuracy
Linear Regression 92%
Lasso Regression 88.37%
Decision Tree 80.80%
Random Forest 90.95%
22
COMPARATIVE STUDY OF ALL THE ALGORITHMS
23
CONCLUSION
Hence out of the four model that we used, the highest accuracy was found to be 92%
which is yielded by Linear Regression Algorithm.
24
REFERENCES
Praful Rane, Deep Pandya and Dhawal Kotak, “Used Car Price Prediction”, International Research Journal of
Engineering and Technology (IRJET),2021.
Pattabiraman Venkatasubbu & Mukkesh Ganesh, “Used Car Price Prediction Using Supervised Learning
Techniques”, International Journal of Engineering and Advanced Technology (IJEAT), 2019.
Ketan Agrahari, Ayush Chaubey ,Mamoor Khan and Manas Srivastava,” Used Car Price Prediction Using Machine
Learning”, International Journal of Innovative Research in Technology(IJIRIT),2021.
https://www.kaggle.com/datasets/balaka18/quikr-cars-scraped
THANK YOU !
