Final viva-voce (12th Sept 2024) for MCA by Research on
PREDICTIVE ANALYSIS FOR BIG MART SALES
By
Mr. SHARATHBABU J Y
USN
4UB22MC090
Under the Guidance of
Dr. MD IRSHAD HUSSAIN B
Assistant Professor
Department of MCA , UBDTCE
Davangere - 577 004
CONTENTS
• ABSTRACT
• INTRODUCTION
• LITERATURE SURVEY
• TECHNOLOGIES USED
• SYSTEM DESIGN
• SCREENSHOTS
• CONCLUSION
• FUTURE ENHANCEMENTS
• REFERENCES
2
ABSTRACT
The Big Mart Sales Prediction project is creating a web application that forecasts
product sales across multiple stores using machine learning in an effort to enhance retail
operations. Retail managers and business analysts can input product and outlet details into this
Flask-built application to receive real-time sales projections. The tool helps optimize inventory
levels, set competitive prices, and make strategic decisions by offering precise projections. This
eventually improves operational efficiency and profitability. The system consists of two modules
a user module that retrieves sales forecasts and an admin module that handles user registrations
and FAQs This project promotes customer happiness and corporate growth by streamlining
inventory management, cutting waste, and improving the overall shopping experience.
3
INTRODUCTION
AIM:
The aim of the Big Mart Sales Prediction project is to develop a machine learning-based web
application that accurately forecasts the sales of various products across different retail outlets.
ABOUT:
Big Mart Sales refers to the commercial activities of a large retail chain that offers a wide
variety of products, ranging from food and beverages to household items and personal care
products. The retail environment is highly competitive, and understanding sales dynamics is
crucial for maintaining profitability and customer satisfaction
4
LITERATURE SURVEY
• A literature review is more than just a list of references; it is a comprehensive
synthesis, analysis, and critical appraisal of earlier works relevant to a particular
topic. It needs to perform the following vital duties: To comprehend the extent and
complexity of data related to the topic, the procedure consists of the subsequent
stages: The process starts with finding relevant books, articles, and other academic
resources. Next, these sources' applicability, authority, and dependability are
evaluated. Third, difficulties that have surfaced, recurring themes, and gaps in the
literature are identified. Fourth, the themes that come out of the sources'
comparison and contrast define the literary structure.
5
EXISTING AND PROPOSED SYSTEM
EXISTING SYSTEM:
To predict sales, the majority of retail organizations currently use simple statistical
models and historical data analysis. These conventional methods frequently rely on
handwritten spreadsheets and don't always take complex variables like shifting
customer behavior or market conditions into consideration. They may therefore be
ineffective and prone to mistakes, which can cause problems with pricing and
inventory control. By utilizing cutting-edge machine learning algorithms to provide
more precise and up-to-date sales projections, the Big Mart Sales Prediction project
aims to enhance these approaches and assist merchants in optimizing their
operations and making wiser decisions.
6
Cont...
PROPOSED SYSTEM:
A web application that uses state-of-the-art machine learning to anticipate sales for
different retail locations is the suggested solution for the Big Mart Sales Prediction
project. The two main parts of the Flask-created application are the admin module,
which handles user registrations and FAQs, and the user module, where retail
managers submit data to get real-time sales estimates
7
TECHNOLOGIES USED
• Visual Studio
• HTML
• CSS
• SQLite3
• Python
8
SOFTWARE AND HARDWARE REQUIREMENTS
Hardware Requirements
• Processor : Intel 13 Processor or higher.
• RAM : 8GB or more.
• HDD : 512 GB or more.
• Keyboard
• Mouse
Software Requirements
• Operating System : Windows 11.
• Front End : Python.
• Back End : SQLLite3.6.
FUNCTION AND NON FUNCTIONAL
REQUIREMENTS
FUCTIONAL REQUIREMENTS:
Admin
Admin can view registered users.
Add can add and view FAQ’s
Admin can view users list.
Users
User can register and login to the application.
User can be able to input product details.
User can view the predicted sales based data.
User can view the FAQ’s list.
10
Conti...
NON FUCTIONAL REQUIREMENTS :
• performance
• security
• usability
• dependability
11
SYSTEM DESIGN
MODLUES USED:
1. Admin Module:
To ensure the smooth and efficient functioning of the system, the admin module offers
extensive control and administration features. Administrators have the ability to
inspect, add, edit, and remove users from user accounts, in addition to keeping an eye
on their activity. In order to better serve users, they can also administer faqs by adding,
changing, and removing entries. This module makes that the system functions
properly, meets user needs.
[Link] Module:
The sales prediction tool can be effectively utilized by retail managers and business
analysts thanks to the user-friendly interface offered by the user module. The system's
functions are easily accessible to users upon registration and login. After logging in,
users may input pertinent product and outlet details to get real-time sales projections,
which enables them to plan their pricing and stocking strategies. 12
USECASE DIAGRAM
Login
User
Add FAQ
View FAQ
Figure 1: Use case diagram for administration module
13
DATA FLOW DIAGRAM
Admin Login
Login Admin
View User
Add FAQ
View FAQ
Figure 5: Data flow diagram for admin
The DFD of the admin explains the logical connection between
the database and the working of the admin in the application.
14
DFD FOR USER
Login
User
Login User
FAQ
ADD
View FAQ
Figure 5. Dataflow diagram for user
15
DFD FOR ADMIN
Login
Admin
Login Admin
View User
Add FAQ
View FAQ
Figure : Data flow diagram intended for administrators
The logical relationship among the database and the admin's operations within the
program is explained in the admin's DFD. 16
SCREENSHOTS
Fig : Admin Login
Conti….
Figure : User Login
18
Conti….
Figure 6.1.11 : Predictive Analysis for Big Mart Sale
19
CONCLUSION
Big Mart Sales Prediction project successfully developed a reliable machine
learning model integrated into a user-friendly web application using the Flask
framework. By leveraging historical sales data and relevant product and outlet
features, the model accurately predicts product sales across various retail outlets.
The application provides separate modules for administrators and users, allowing
administrators to manage FAQs and view user information, while users can
register, log in, input product details, and receive real-time sales predictions.
Through this project, retail businesses can make informed decisions to optimize
inventory management, pricing strategies, and overall operations, ultimately
leading to enhanced profitability and competitiveness in the retail market.
Continued maintenance and updates will ensure the application remains effective in
meeting the evolving needs of retail managers and business analysts.
20
FUTURE ENHANCEMENTS
Advanced reporting and analytics: include more extensive reporting and analytics tools to give more in-depth
understanding of performance indicators and sales patterns.
Mobile application: create a mobile version of the application so customers may access other features and
sales forecasts while they're on the go.
Real-time data integration: to improve sales forecast accuracy and react fast to market shifts, incorporate real-
time data streams.
Ai-driven insights: apply ai-driven suggestions and insights for predictive analytics-based inventory, price,
and promotion optimization.
Multi-language support: to reach a wider audience and enhance accessibility, include multi-language support.
21
REFRENCES
• The complete reference python: By Martin C. Brown .
• Python machine learning: By Sebastian Raschka & Vahid Mirjalili .
Referred website
[Link]
[Link]
[Link]
ig_Mart_Sales_Prediction
[Link]
378532586_Big_Mart_Sales_Prediction_Using_Machine_Learning
22
Thank you
23
