I am a business analytics & economics student who is passionate about programming and data science. I am actively looking for opportunities to grow my skillset in statistics, coding, data structures and algorithms and deep learning.
| Project | Description | Website/Code |
|---|---|---|
| Stock News Sentiment | Led a team in NUS Fintech Society to create a website that provides news sentiment scores for NYSE and NASDAQ stocks for investors and traders to make informed decisions. Explored using news sentiment scores for predicting stock performance over a short-medium term time period. | Website | Code |
| Open Banking Project | Led a team in NUS Fintech Society to create a one-stop-shop FX platform that not only provides investors and businesses with the latest FX rates and financial, but also help them make better Forex transaction decisions by providing future price movement signals and sentiment of news headlines with machine learning models. | Website | Code |
| Insurtech Computer Vision Project | Led a team in NUS Fintech Society to develop computer vision models, such as Mask-RCNN and YOLOv5, that detect different types of car damages and provide an estimate of the cost of repairs. I showcased the project at the Singapore FinTech Festival, Talent Pavilion 2021, where I also introduced participants to computer vision techniques and shared quick start guides for them to embark on their own projects. Deployment on Heroku was done using Docker to overcome the 500MB file size restriction set by Heroku. |
Website | Code | Sample Images |
| AWS Build On, ASEAN 2021 - Team Retype | Developed NLP models with BERT, DistilBert and AWS Comprehend to conduct sentiment analysis of chat messages and give each user a Friendliness score that could supplement buyer/seller reviews on C2C platforms. In addition, created a lowball filter function that redirects frivolous offers into a separate chat folder so that sellers can focus on legitimate offers. Won the first place in the AWS Build On, ASEAN 2021 competition for the challenge statements by Carousell. |
Website |
| JustNice Recipe App | An easy-to-use online platform to explore and store recipes. What’s more, create grocery lists with your recipes to help plan your next trip to the supermarket. | Website | Code |
| News Headlines Sentiment Analysis | Employed transfer learning with BERT and RoBERTA to perform sentiment analysis on news headlines. Achieved first place in the Kaggle Competition organised by NUS FinTech Society with an accuracy of 90% | Code |
| Stock Prediction | Used ARIMA model to predict stock prices for a Kaggle Competition organised by NUS FinTech Society | Code |
- Python
- C++
- Java
- JavaScript
- HTML
- CSS
- R Programming
- SQL
- sh