Market Basket Analysis using
Python
Market Basket Analysis using Python is a project focused on uncovering shopping patterns within retail transaction
data. It utilizes Python's data science toolkit to extract, process, and analyze data, uncovering associations between
different products bought together.
Ashit Singh (MB23047)
Utsav Seth (MB230)
Sunil Choudhary (MB23004)
Anurag Mishra (MB230)
Introduction
Retail Landscape Data-Driven Decisions Python Toolkit
The report introduces the concept It emphasizes the project's It utilizes Python's data science
of Market Basket Analysis, a relevance in today's data-driven toolkit to extract, process, and
technique widely used in retail for decision-making landscape in analyze data, uncovering
understanding customer purchase retail. associations between different
behaviour. It emphasizes the products bought together.
project's relevance in today's data-
driven decision-making landscape
in retail.
Objective
Enhanced Strategy Targeted Marketing Optimized Product
Placements
The main goal is to analyze Derive actionable insights for
transactional data to identify business strategy
frequent item combinations enhancements, such as Identify frequent item
and derive actionable insights targeted marketing and combinations and derive
for business strategy optimized product actionable insights for
enhancements, such as placements. optimized product
targeted marketing and placements.
optimized product
placements.
Data Collection and Preprocessing
1 Data Gathering
This section would cover how the data was gathered, its characteristics, and preprocessing steps
like cleaning and transformation for analysis readiness.
2 Data Characteristics
Cover how the data was gathered, its characteristics, and preprocessing steps like cleaning and
transformation for analysis readiness.
3 Preprocessing Steps
Cover how the data was gathered, its characteristics, and preprocessing steps like cleaning and
transformation for analysis readiness.
Methodology
Apriori Algorithm Association Rules Data Science Toolkit
Utilization of the Apriori Generation of association rules Python libraries (Pandas, Plotly,
algorithm for finding frequent to understand product affinities. etc.) used in the process.
item sets.
Python libraries (Pandas, Plotly,
Python libraries (Pandas, Plotly, etc.) used in the process.
etc.) used in the process.
Analysis and Findings
1 Antecedents 2 Consequents
These are the items that are considered These are the items that tend to be
as the starting point or “if” part of the purchased along with the antecedents or
association rule. the “then” part of the association rule.
3 Support
Support measures how frequently a particular combination of items appears in the dataset.
Patterns and Trends
Specific Findings Frequent Combinations Customer Behavior
Patterns and trends identified Insights from the association
from the transaction data. The most frequently purchased rules regarding customer
item combinations. purchasing behavior.
Visualizations
Graphical Histograms Graphical Insights
Representations
Descriptions of histograms Other graphical
A discussion of the used in the notebook. representations used in the
visualizations generated using notebook.
Plotly, explaining how they
contribute to understanding
the data.
Conclusion and Business Implications
1 Support and Confidence
The relationship between 'Apples' and 'Bread' shows a support of 0.045752, indicating these items
are frequently bought together.
The confidence levels in both directions are around 28% and 30%, suggesting a moderate
likelihood that purchasing one will result in purchasing the other.
2 Lift Values
The lift values exceeding 1 imply a strong positive association between these items.
Similar patterns are observed with other item pairs like 'Apples' and 'Butter', 'Apples' and 'Cereal',
and 'Apples' and 'Cheese'.
3 Strategic Decision-Making
These findings are crucial for strategic decision-making in product placement and promotional
activities.
