Marketing Campaign Analysis (FB vs AdWords)

(Google Colab, Hypothesis Testing, Regression & Time Series Analysis)

Accomplished a 76% prediction accuracy in ad conversion rates, by applying Linear Regression, Random Forest, and Ensemble Learning to analyze and optimize marketing campaign data.

Enhanced campaign insights with a 0.87 correlation coefficient between Facebook ad clicks and conversions, demonstrating a strong linear relationship that supports data-driven decisions for ad spend and click strategies.

Improved ROI and cost-effectiveness of ad campaigns, by conducting hypothesis testing that demonstrated Facebook’s superior conversion rates over AdWords, leading to more strategic budget allocation.

Demonstrated consistent engagement throughout the week, by analyzing conversion rates across weekdays and finding that Mondays and Tuesdays exhibit the highest conversion rates, informing day-specific marketing strategies.

Identified cost-effective advertising periods, by analyzing CPC trends and recognizing lower CPC months (May and November) as opportunities for maximizing budget efficiency and return on investment.

Pilot Program For Churn Model

(Big Query, Python and Looker Studio)

Led the creation of an AutoML model for a pilot program predicting employee churn. Leveraged Google Cloud's BigQuery for data integration and manipulation, handling 15K+ data rows to enhance model accuracy.

Utilized BigQuery in Google Cloud Platform to merge and manipulate data from multiple sources. Employed IPython notebooks to connect with BigQuery, applying Pycaret for model training, achieving 99% accuracy with Random Forest model selection.

Developed dashboards using Looker Studio that included custom KPIs and visualizations. Identified satisfaction level as the most crucial factor, highlighting technical and support departments with the highest churn rates. Overall churn rate revealed to be 7%, informing strategic HR decisions.

Maximizing Revenue For Drivers

(Statistics and Hypothesis in Python, Google Colab)

Employed statistical analysis and hypothesis testing in Python to identify revenue-maximizing strategies for taxi drivers. Conducted data preprocessing steps, including filtering duplicates and removing null values, to ensure data accuracy.

Utilized stacked horizontal bar charts and pie charts to analyze the ratio between primary payment methods which is Card and Cash (67.3 & 32.7)%. Enabled visualization of payment method distribution, aiding in identifying dominant payment channels and potential areas for improvement.

Applied filtering techniques to remove duplicate values and handled null values effectively. Ensured data integrity and accuracy, laying a solid foundation for meaningful statistical analysis and hypothesis testing.

Famous Painting Analysis

(SQL Project, PG Admin, MS Excel, Data Manipulation)

Utilized SQL queries, including window functions, group by, and joins, to extract insights on artworks, artists, museums, pricing trends, and geographical distributions, informing strategic decisions.

Rediscovered dominant painting styles like Impressionism and identified prolific artists like Renoir, Monet, and Van Gogh, providing insights into art market dynamics and preferences.

Investigated pricing differentials, consumer behavior, and museum opening hours, revealing key findings: 20+ insights, including pricing trends, market demand insights, and accessibility commitments.

Customer Churn Analysis

(Power BI, MS Excel, DAX, Data Transform, Visualization)

Spearheaded the analysis of Customer churn using Power BI. Transformed data, established reference tables, and implemented DAX for visualization, uncovering insights to inform strategic decisions.

Developed pie charts for customer segmentation and line/bar plots for churn rate analysis, leveraging Power BI's capabilities. Enhanced understanding of data patterns, facilitating informed decision-making for retention strategies.

Discovered gender distribution (44% female, 56% male) among churned customers, activity status (52% active, 48% inactive), credit card usage (30% none, 70% with), and country-specific churn rates (24% Spain, 25% Germany, 51% France) Insights for retention strategies.

Coffee Sales Analysis

(MS Excel, Pivot Table, Data Modeling, Data Visualization)

Utilized SQL queries, including window functions, group by, and joins, to extract insights on artworks, artists, museums, pricing trends, and geographical distributions, informing strategic decisions.

Rediscovered dominant painting styles like Impressionism and identified prolific artists like Renoir, Monet, and Van Gogh, providing insights into art market dynamics and preferences.

Investigated pricing differentials, consumer behavior, and museum opening hours, revealing key findings: 20+ insights, including pricing trends, market demand insights, and accessibility commitments.