Projects

CycleSync Dashboard

Explore the urban landscape of Vancouver through the lens of CycleSync, a comprehensive Bikeshare Dashboard that delves into the intricacies of Mobi by Rogers' bikeshare system. This project unfolds vital insights into ride statistics, station activities, and usage patterns, offering a clear and informative overview of the bikeshare ecosystem. From an interactive map showcasing active stations to in-depth analyses of daily trends and environmental influences, CycleSync serves as a valuable tool for understanding and studying urban mobility dynamics.

Click here to learn more about the project!

DashKick Analytics

Link: GitHub Repository

A comprehensive platform for football enthusiasts to analyze the 2023 Premier League season. By leveraging data from the API-Football, the DashKick Analytics package aims to unravel the stories within the numbers, predict outcomes, and visually represent the pulse of each match and team.

Click here to learn more about the project!

Predicting Milk Quality

Link 1: Project Slide Deck

Link 2: Project Report

A classification study to enhance milk quality assessment through cutting-edge supervised machine learning. This project employs Python and scikit-learn to predict milk grades. Key models, including K-Nearest Neighbors and Random Forest, showcase robust performance, with pH emerging as a crucial predictor. The project employs techniques like SMOTE for class imbalance mitigation, and results in commendable recall rates, essential for precision in milk quality predictions.

Dive deeper into the methodology, findings, and implications in the detailed breakdown here.

Data Analytics Case Study

Link to project slide deck: Datathon Slide Deck

Link to award certificate: BOLT Award Certificate

This project was the final deliverable that my team presented during the 2022 BOLT x Accenture Data Analytics Bootcamp. We were placed 4th (out of 30 teams) for our presentation!

For this project, a machine learning model was devised to identify and analyze the most important predictors of employee attrition at Summit Biotech. After the top issues were identified, a set of data-driven recommendations on how the company can best address the needs of the identified subset and decrease the turnover rate were formulated.

Minitab: Statistical Experimental Design Techniques

Link to project report: FNH 403 Final Report

As part of my FNH 403 final project, my team created a systematic plan to develop vegetable crackers made up of Brewer's Spent Grain as one of its main ingredients. We utilized several statistical design of experiment (DOE) methods, such as:

Fractional factorial design -> to determine optimal ingredient amounts (initial screening)
Mixture design (extreme vertices) -> to create a flour mixture formulation with optimum nutrient content, colour and flavour
Central composite design -> to find the perfect parameter combination for an optimum flavour, crispness, and appearance