Projects
In this project, I built a predictive model to forecast YouTube video views by analyzing the sentiment of 100 comments collected from almost 24000 videos via the YouTube API. Using NLP techniques, I classified comments into sentiment categories and applied machine learning models like Random Forest, Ridge Regression, and K-Nearest Neighbors, achieving 85% accuracy. By correlating sentiment with engagement metrics such as likes, dislikes, and comment volume, I developed a model that effectively predicts video viewership.
Created an interactive and informative dashboard to analyze the reviews of British Airlines with the option of picking the desired available metrics and other filters (6 different types) if there is to be understood about specific data points along with the visualizations (4 different visualizations including geographical) being a dynamic filters themselves.
In this project, I conducted an extensive analysis of layoffs data using SQL, working with around 10,000 records to uncover patterns and trends across industries, companies, and regions. I implemented data cleaning techniques, including deduplication, null value handling, and data standardization, reducing data inconsistencies by 20%. Through exploratory data analysis (EDA), I identified key trends such as layoffs by company, industry, and time, and performed aggregations to analyze the impact of various factors on workforce reductions.
Developed a machine learning model to predict heart disease based on a medical dataset of 300+ patient records. Using algorithms such as K-Nearest Neighbors, Decision Trees, and Logistic Regression, I achieved a prediction accuracy of 90%. I applied extensive feature engineering and data visualization techniques to identify key risk factors, optimizing the model for better performance.
Created an interactive Excel dashboard to analyze Dmart's 2024 sales data, providing a comprehensive overview of category-wise, sub-category-wise, segment-wise, and region-wise sales performance. The dashboard features dynamic slicers for state and sub-category, as well as a ship date timeline, allowing for in-depth analysis of key sales trends and patterns. By visualizing metrics such as sales by ShipMode and customer segment, I enabled actionable insights into business performance, enhancing data-driven decision-making for inventory management and sales strategy.
Performed an exploratory data analysis (EDA) on PlayStation 5 game data to uncover insights into game sales, genre popularity, and customer preferences. Using Python for data cleaning and visualization, I analyzed key metrics such as game sales performance, release trends, and user ratings. By identifying correlations between game genres and sales success, as well as trends in release dates and customer engagement, the analysis provided valuable insights into market dynamics, helping optimize decision-making for game developers and marketers.
