University College London | Sep.2025-Sep.2026

- Machine Learning for Data Science, Spatial-Temporal Data Analysis and Data Mining, Spatial Databases and Data Management, Geospatial Programming, Spatial Analysis and Geocomputation, Sensors and Location, Research Project

Soochow University | Sep.2021-Jul.2025

- GPA: 3.8/4
- Modules: Multivariate Analysis, Data Mining, Machine Learning, Big Data Processing, Data Engineering, Artificial Intelligence.

Programming based | 2020 -

Site selection with MaxEnt
Maximum Entropy (MaxEnt) model to conduct a multi-criteria spatial analysis of data centre site selection across England.

Rail Station Spatial Decision Support System (SDSS)
Introduces a multi-level spatial decision support system for Interurban Passenger Rail Transport (SIC 49.1). Integration of Spatial SQL (PostGIS) with operational and financial data enables automated facility management decisions, from targeted gate repairs to comprehensive station-wide budget allocation.
Topological Network Resilience & Structural Heterogeneity
Investigating the structural vulnerability of traffic networks by integrating TDA with Social Network Analysis (SNA).

Multiscale Spatial Discovery & Bayesian Modelling
Developed a hybrid framework combining Spatial Autocorrelation (Moran's I, Getis-Ord Gi*) and Bayesian Networks to identify high risk road factor combinations.

Visualizing High-Dimensional Data with Topological Data Analysis: A Case Study on Traffic Accident Risk | 2023/11 ~ - Submitted to Special Issue on "Advances in Data Analysis and Visualization" (Electronics)
- White box non-linear dimensional reduction methods (Mapper Algorithm) were used to reveal the real traffic collision structure in Taiwan.
- Applied the Mapper algorithm to separate complex accident heterogeneity into distinct clusters, thereby improving the precision and stability of fatality risk identification.

Traffic Accident Severity Classification | 2023/11 ~ - International Conference and Annual Meeting of the Chinese Institute of Transportation
- Employs Topological Data Analysis to analyze the severity of traffic accidents
- Captures the latent structure in road design, weather, and passenger behavior data.

MRT station crowding prediction model: application based on machine learning methods | 2023/11 ~ - International Conference and Annual Meeting of the Chinese Institute of Transportation
- Combines LSTM for analysing temporal MRT OD data and ConvLSTM for capturing spatial YouBike patterns to accurately predict station crowding.

Office of Institutional Research | Feb.2025-Jul.2025 Contract-based Staff

School database inventory and compilation, processing school affairs data through information retrieval and data analysis

Constructing Demand-Responsive Transportation Systems | Sep.2024-Jun.2025 Contract-based Project Researcher

Tourism flow prediction based on data mining and machine learning methods

Trading Education Platform (Startup) | Jun.2021-Nov.2023 Blockchain Engineer

- Built a risk-free trading training platform that helps in the real market
- Deployed multiple smart contracts for users to interact while learning trading skills
- With an average of around 600 users per year and a total of 16,000 registered users

TDA:Mapper Algorithm
This package is an implementation of the Mapper algorithm in R, speeding up the process of creating a topological map of high-dimensional data, and provide a playground for users to experiment with the algorithm. Several clustering methods and plots are provided also.

TDA:Simplicial Complex
Provide a user-friendly simplicial complex application that is compatible with the wide range of statistical methods available in R.

Hackathon | 2020 -

2021 Backyard Hackathon: Old Street Digitalization Project
(Gold medal with 100,000 TWD) Create a platform for locals to design routes for users, allowing them to collect region-specific commemorative NFTs at designated locations. Additionally, incentivize users to gather litter in exchange for discount vouchers. Team: Stetri.

2022 BNP PARIBAS CARDIF: Health Tracking
(Gold medal with 60,000 TWD) Encouraging Regular Exercise Through Wellness Insurance Policies Integrated with Google Fit for Program Analysis and User Health Tracking.

2023 Innovation Hackathon Competition: Deepfake classification
(Honorable Mention with 2,000 TWD) A website including Image/Video classification and GAN to generate an image, the concept behind establishing this website is to prevent individuals from being deceived by AI-generated deepfake manipulations. Also addes GA4 for analytics and MySQL for data storing.