Data Science Portfolio

This repository contains my data science portfolio, showcasing projects across various domains, including food order analysis, EdTech industry analysis, and a recommendation system for Amazon products. Each project demonstrates my skills in data preprocessing, exploratory data analysis (EDA), data visualization, statistical analysis, and machine learning.

Projects

1. Food Order Analysis with Google Colab

This project involves analyzing a dataset from a food delivery platform to derive insights into customer ordering patterns and business performance. The analysis covers revenue trends, order frequency, and high-value customer identification.

Key Objectives

• Identify the most frequently ordered items.
• Analyze peak order times and revenue trends.
• Determine high-value customers based on order frequency and total spend.

Process Overview

• Data Loading: Mounted Google Drive and loaded the dataset into a Pandas DataFrame.
• Data Cleaning: Handled missing values and corrected data inconsistencies.
• EDA: Generated summary statistics, visualized order distributions, and revenue trends.
• Visualization: Used Matplotlib and Seaborn to create bar charts, line plots, and histograms.

Technologies Used

• Python (Pandas, Matplotlib, Seaborn)
• Google Colab

2. EdTech Industry Hypothesis Testing

This project focuses on statistical analysis and hypothesis testing in the EdTech sector. The goal is to validate key business hypotheses about user engagement and course completion rates, providing data-driven insights to improve user experience.

Key Objectives

• Test whether course completion rates differ significantly across user segments.
• Analyze the relationship between engagement metrics and user retention.

Process Overview

• Hypothesis Formulation: Defined null and alternative hypotheses for various business questions.
• Statistical Testing: Conducted t-tests and ANOVA to validate hypotheses.
• Visualization: Presented results using box plots, histograms, and scatter plots.

Technologies Used

• Python (SciPy, Statsmodels, Seaborn)
• Jupyter Notebook

3. Amazon Product Recommendation System

This project involves building a recommendation system for Amazon products using collaborative filtering and content-based filtering approaches. The aim is to suggest relevant products to users based on their past interactions and product attributes.

Key Objectives

• Build a collaborative filtering model using user-item interaction data.
• Implement a content-based recommender using product descriptions and metadata.
• Evaluate the performance of the recommendation system using appropriate metrics.

Process Overview

• Data Preprocessing: Cleaned and prepared the user-item interaction matrix.
• Model Development: Implemented collaborative filtering using matrix factorization techniques.
• Content-Based Filtering: Extracted features from product metadata and computed similarity scores.
• Evaluation: Measured the accuracy of recommendations using precision, recall, and F1-score.

Technologies Used

• Python (Pandas, NumPy, Scikit-Learn)
• Jupyter Notebook

Technologies Used Across Projects

• Python (Pandas, NumPy, Matplotlib, Seaborn, SciPy, Scikit-Learn)
• Jupyter Notebook / Google Colab