Portfolio

EDA - Exploratory Data Analysis

Click the project titles to view the full descriptions.

Market Basket Analysis

Implemented Market Basket Analysis: Analyzed customer purchase patterns in an online retail dataset to identify frequently bought-together items. This involved data cleaning, binary encoding for presence/absence of products, and the Apriori algorithm to generate association rules.

Generated Actionable Insights: Identified strong co-occurrence patterns between products and provided recommendations for improved product placement, inventory management, and potential bundled promotions to increase sales.

Retail Sales EDA in Python

Performed Customer segmentation by age and gender, identifying variations in daily and monthly spending patterns across product categories.

Analyzed customer demographics (age, gender) to understand their purchasing behaviour and identify key product categories for different customer segments.

Tech Layoffs Data Cleaning in SQL

Cleaned and prepared the Tech Layoffs 2020-2024 dataset using SQL (sqlite3) by addressing missing values, spelling mistakes and duplicate rows.

Tech Layoffs Tableau Dashboard

Used the cleaned dataset above to create an Interactive Tableau Dashboard.

Mental Health EDA in Python

Used Pandas, as well as Seaborn and Matplotlib to investigate relationships between stress levels, mental health history, and treatment seeking among students, segmented by country and gender.

Unveiled patterns in mental health data by analyzing factors like gender distribution across countries, occupations, and treatment-seeking behaviour.

Retail Sales EDA in SQL

Leveraged SQL (SQLite3) to analyze retail sales data, extracting insights on customer demographics, purchasing behaviour by product category, and price variations.

Shared a tutorial of how to use SQLite3 in Jupyter Notebook.

Spotify data 2023 EDA

Cleaned and prepared the Spotify dataset for analysis, addressing missing values and irrelevant features.

Dived deep into the top 20 most-streamed tracks, exploring artist prevalence and audio feature co-occurrence (e.g., examining if high-energy songs tend to be danceable).

Investigated how musical key and mode have evolved over time.

I will be exploring and sharing more projects on Kaggle linked to Github