Semi-Supervised Variational Autoencoder and Classifier for Synthetic Data Generation

This project uses VAEs and CVAEs in PyTorch to generate synthetic, class-conditional data in a semi-supervised setting. It evaluates data quality with Inception Score and FID, and tests generalization with an SVM classifier. PCA and t-SNE visualize class distribution for model insights.

Marketing-Ad-Campaign-Data-Analysis

This project aims to analyze and track key performance metrics for two platforms used to process orders: Shopify and Checkout Champ (CoC). The objective is to generate various business metrics, including order count, revenue, unique customers, and more, both for each platform.

Ecommerce Order Tracking Analysis using Python and Excel

This project addresses an order-tracking issue on our ecommerce platform. The goal is to determine the volume of orders affected by tracking issues compared to the total number of orders.

Interactive Story Generator with WebSocket and GPT-2

his project is a dynamic, user-driven storytelling experience powered by GPT-2 and WebSockets. Users make choices that shape the narrative, with real-time interaction handled by a WebSocket server. The story is narrated using text-to-speech (pyttsx3) to enhance immersion.

Temporal Knowledge Graph of India-Pakistan Relations using Python, NetworkX, and spaCy

This project creates a temporal knowledge graph using Python, NetworkX, and spaCy to trace India-Pakistan relations over time. It extracts and links entities from historical data, mapping conflicts, alliances, and events.

Tableau: Korea’s Suicide Rates by City and County

I developed a Tableau dashboard featuring line charts to analyze suicide rates in Korea by city and county. The data trends over time revealed key regional patterns. This visualization helps in tracking changes and identifying areas of concern.

SQL & Python: Quantative Marketing Analyisis of Sales Data

I used SQL, and Python to analyze Carvana's marketing data, focusing on user click data and sales attribution. By querying and processing data, I calculated key metrics such as Customer Acquisition Cost (CAC) and Return on Investment (ROI) for each marketing channel.

Cisco: Excel tool for calculating GHG Carbon Emissions Generated by Teleworking

I led a team of four to develop an advanced Excel tool for measuring Cisco's greenhouse gas emissions. This project, undertaken under the guidance of Cisco's Chief Sustainability Officer, enhanced the accuracy and efficiency of emission tracking and reporting.

Emotion-Based Music Harmonizer Python App.

This project provides an intuitive tool for mixing audio tracks to match a chosen emotion — calm, sad, or energetic. Using Python, pydub, and tkinter, the harmonizer adjusts melody, harmony, bass, and percussion layers based on emotion-specific filters, tempo, and volume effects.

Machine Learning: Predicting Red Wine Quality

I analyzed and predicted red wine quality using a dataset of physicochemical properties. Using Python and Scikit-learn, I conducted data analysis, preprocessing, and model evaluation. I trained and compared models to find the most accurate predictor.

Power BI: Top 50 Songs on Spotify in 2019

I created a Power BI dashboard for the Top 50 Songs on Spotify in 2019, analyzing danceability, beats per minute, energy, and genre. This interactive tool highlights trends in music characteristics across genres.

Churn Rate Prediction for Cell Phone Customers

The Cell Phone Churn project predicts customer churn using machine learning models and data analytics. Random Forest was optimized through hyperparameter tuning, improving key metrics like accuracy and recall. Top churn factors were identified for better decision-making.

Netflix Movie Recommendation System

This project creates a movie recommendation system using K-Nearest Neighbours (KNN), Alternating Least Squares (ALS), and Singular Value Decomposition (SVD). It leverages collaborative filtering to provide personalized movie suggestions based on user interactions.

Mlops: House Rent Prediction App

I led a House Rental Prediction project using MLOps principles to automate the entire model lifecycle. By integrating AWS for cloud deployment and Jenkins for CI/CD, I streamlined the pipeline from data ingestion to model deployment and monitoring. This ensured efficient, scalable, and automated pre

R Programming: Mall Customer Segmentation using R

I used R programming to segment mall customers based on age, income, and spending habits using K-Means clustering. I applied techniques like the Elbow and Silhouette methods to determine the optimal number of clusters. This analysis helped me create key visualizations and provide insights for person

SAS Health Metrics Analysis: BMI Categorization and Gender Insights

The project analyzes individual health metrics by calculating BMI and categorizing health statuses. Using SAS programming, it visualizes data through pie charts and generates frequency tables to explore the relationship between gender and BMI.