Projects

Email Spam Classification with Custom Naive Bayes & k-NN Algorithms

This project implements custom Naive Bayes and k-Nearest Neighbors classifiers from scratch to identify email spam, using core Python libraries (NumPy, Pandas). The Naive Bayes model incorporates class priors, likelihoods, and Laplace smoothing, while the k-NN classifier optimizes prediction accuracy through a custom distance metric. The project emphasizes hands-on understanding of machine learning fundamentals and is designed to process both text and numeric datasets. Ideal for showcasing proficiency in data pre-processing, model building, and evaluation techniques.

NYC Airbnb Market Analysis

Analyzed the NYC Airbnb market to uncover high-yield investment opportunities using advanced data cleaning, occupancy rate calculations, and revenue estimation techniques. Delivered actionable insights and a professional summary to guide strategic real estate investments, showcasing expertise in turning raw data into impactful decisions.

Optimizing Ride-Sharing Strategies: Data-Driven Insights for Market Entry Success

Conducted an in-depth analysis of Chicago's ride-sharing market for Zuber using SQL, uncovering passenger trends, neighborhood preferences, and the impact of weather on ride frequency and duration. Delivered actionable insights to inform strategic decisions, showcasing expertise in leveraging data to optimize service efficiency and customer satisfaction.

Driving E-Commerce Success: Conversion Funnel and Retention Analysis

This project dissects user activity on an e-commerce platform to optimize customer engagement and retention. By analyzing conversion funnels, performing cohort-based retention tracking, and calculating retention rates, it identifies key drop-off points and trends. The insights guide actionable strategies to enhance user experience, improve conversion rates, and foster long-term customer loyalty.

Superstore Profitability Analysis: A Data-Driven Approach to Financial Stability

Discover an in-depth analysis of Superstore's performance, designed to uncover key profit drivers and mitigate loss-making areas. This project highlights actionable insights, including profit centers, high-return advertising strategies, and loss-reduction techniques. By leveraging advanced data visualization and analysis, it provides recommendations to optimize operations, improve return on ad spend, and guide strategic decision-making for long-term profitability.

Reducing Returns: Data-Driven Solutions for Superstore Profitability

This project leverages Tableau to analyze Superstore's return trends across regions, product categories, and seasons. By identifying high-return drivers such as specific products and geographic hotspots, the analysis provides actionable recommendations to reduce returns, optimize sales strategies, and improve profitability.

Shopify App Analysis: Insights into App Success and User Engagement

Explore the key drivers of success for Shopify apps through an in-depth analysis of user reviews, ratings, and developer responsiveness. This project leverages Power BI to create dynamic visualizations, revealing trends in app performance, review dynamics, and the impact of developer engagement. With actionable insights, this analysis empowers app developers and stakeholders to make data-driven decisions that enhance user satisfaction and app profitability.

Zomato Customer Segmentation: Insights for Targeted Marketing

Performed demographic and behavioral segmentation of Zomato's users by analyzing customer orders, spending habits, and cuisine preferences. Delivered actionable insights through Tableau visualizations, enabling targeted marketing strategies and enhancing user engagement.

Student Performance Analysis: Insights into Academic Success

Analyzed factors influencing academic outcomes, including attendance, study habits, and socioeconomic conditions. Developed an interactive Tableau dashboard to explore correlations between resources, family income, and student performance, providing actionable insights to optimize educational strategies and support systems.

Custom Data Analysis and Visualization: Unlocking Insights with Python

Developed a robust framework for statistical analysis and data visualization using custom Python classes. The project handles data preprocessing, computes key metrics (mean, range, variance, etc.), and generates intuitive visualizations such as scatter plots and pair plots to uncover relationships and trends. Applicable across domains like finance, healthcare, and engineering, it empowers data-driven decision-making with efficient data handling and advanced statistical techniques.

Principal Component Analysis (PCA) with Covariance Matrix Approach

This repository contains a high-performance Python library, PCA_COV, designed for performing Principal Component Analysis (PCA) using the covariance matrix method. Ideal for data analysts and scientists, it simplifies dimensionality reduction, enhances data visualization, and supports exploratory data analysis. Key features include variance analysis, data normalization, elbow plot generation, and dimensionality reduction, all aimed at transforming complex datasets into actionable insights for machine learning and statistical modeling.

Custom Radial Basis Function (RBF) Neural Network for Function Approximation & Pattern Recognition

A specialized implementation of an RBF neural network utilizing Gaussian kernels and K-means clustering for efficient nonlinear data modeling. Ideal for regression, classification, and signal processing tasks, this project demonstrates how RBF networks can be applied to solve complex problems, including handwritten digit recognition with 80% accuracy on the MNIST dataset. With a focus on modularity and performance, this network showcases skills in Python, NumPy, machine learning, and data visualization.

Custom K-Means Clustering Algorithm for Data Analysis & Image Compression

Developed a flexible, Python-based K-Means clustering implementation for unsupervised learning tasks, including data clustering and image compression. Utilized libraries like NumPy, Pandas, and Matplotlib for data manipulation and visualization. Demonstrated practical applications in reducing image complexity and optimizing clustering processes for real-world datasets.

Music Streaming Data Analysis

This project analyzes Fresh Beats, an innovative music streaming platform focused on promoting emerging artists and delivering engaging user experiences. Through comprehensive data analysis, it uncovers insights into user behavior, genre popularity, and subscription trends. Key recommendations include strategies for enhancing user retention, boosting artist visibility, and optimizing marketing efforts to drive revenue growth.