Portfolio

---

Data Science Projects

ParkFinder

Real-time query, tracking, and statistical analysis of available parking spaces in Taipei City

Finding available parking spaces in Taipei City is always a hassle, especially in unfamiliar areas. While the government provides real-time information through the Parking Information System, it lacks accumulated historical data for users to anticipate future parking availability. To address this issue, my friends and I have designed a website that retrieves parking space information from the system. We aggregate the available parking slots for fixed time periods at various parking lots and offer interactive visualization tools to provide user a intuitive understanding of parking availability trends.

---

Sports Betting Freerider

Elevate the Betting Game with Data-Backed Strategies

Sports betting represents a unique form of lottery. Unlike conventional lotteries that are determined purely by chance, it offers the potential for long-term profitability through targeted research. However, we prefer profits without the burden of research, don’t we? In light of this, I’ve devised a web scraping system to capture the prediction results of top-ranking players from the Playsport. This system collates their wagers, identifies the most favored choices, and provides users with effortless access to potential profitable betting directions. It encompasses web data scraping, data aggregation, and delivery to user mailboxes, utilizing Google Sheets as a compact database for accessing results, and conducting statistical analyses of prediction accuracy.

---

Stock Price Movements Prediction

Forecasting Market Moves with Textual Analysis and Machine Learning

News analysis has always been an important aspect of stock investment, but it couldn’t be systematically utilized by investors due to the difficulty in quantifying information. With the advancements in text mining and machine learning, such analytical methods have re-emerged with fresh perspectives, as highlighted in On the Importance of Text Analysis for Stock Price Prediction. This project, which is a final program for Machine Learning and Econometrics, aims to predict short-term fluctuations in Taiwanese stocks using various machine learning algorithms. The textual data include news articles and forum posts on PTT, which are provided by the professor. Although the training results were not sufficient to serve as practical investment strategies because of the limited amount of data, this project still provided me with valuable practice opportunities.

---

Economics Research

The Effects of Parental Death on Offspring’s Marital Decisions: Evidence from Taiwan

A Pioneering Study in Taiwan Reveals the Potential Effect of Mother-in-law problem using Event Study Method

---

How does National Pension Reform affect household economic behaviors? Evidence from Taiwan

Unveiling Household Responses to Taiwan’s National Pension Reform with Difference-in-Differences method

---

What is the cause and effect of the domestic division of unpaid labor? Evidence From Taiwanese Family

Decoding Household Dynamics Challenging Conventional Wisdom on Unpaid Labor in Taiwan with Ordered Probit model

---