Projects
A selection of projects that I've worked on
Unit Ministry Team Automated Email Dispatch Web Application
Designed and built a web application using Flask and frontend languages, following 3-tier architecture principles to provide an intuitive interface for event organizers. Utilized Google Forms to collect registrations, automatically storing data on the Google Cloud Platform. Used OAuth and the Google Sheets API to securely extract data into a pandas DataFrame for flexible manipulation and analysis. Automated personalized email communication to participants using the Python SMTP library with TLS encryption, improving data security. Enhanced participant experience and engagement by streamlining manual processes and ensuring efficient information flow. Technologies: Google Cloud, pandas, Flask, HTML, CSS, Javascript, API, OAuth
Analyzing the Impact of Audio Features, Genre, and Charting Metrics on Song Popularity: A Data Driven Study on Spotify Data
Conducted descriptive statistics and inferential analysis on Spotify data to identify correlations and differences between audio features and song popularity on the Spotify charts. Utilized linear and logistic regression models to identify significant predictors of the popularity score of a song, including danceability, energy, loudness, and genre. Provided insights and potential growth strategies for music artists, record labels, and streaming services to optimize their content and engage their users based on the findings from the analysis.
Facial Recognition using Principal Component Analysis (PCA)
Developed a facial recognition system using principal component analysis (PCA) for dimensionality reduction and achieved an accuracy of over 90%. Implemented eigenvectors/eigenfaces analysis to identify the most important features of facial images and reduce the number of dimensions in the dataset. Analyzed the principal component face to gain insights into the underlying structure of the facial features and improve the performance of the facial recognition system.
SPX Prediction using Macro-economic Indicators
Extracted, cleaned, and analyzed 15+ economic datasets drawn from TradingView to search for indicators leading the stock market. Created Pearson-correlation Heatmap to identify correlations and plots to confirm trends. Conducted ADF test and Granger Causality test for stationary of data and hypothesis testing, then applied ARIMA autoregression model to make forecasts, fitting the data into Linear Regression ML model (~60% accuracy in prediction).
Business Analytics I Academic Project
Developed a comprehensive Tableau Story with 3 visualization worksheets (bar chart, tree map, geographic heatmap) by analyzing Chicago school dataset in .csv format. Identified positive correlation with college admission in categories of school type, average SAT/ACT score, school district location, and school rating.
