Objective:
This project integrates the distribution of common hazardous species in Taiwan’s mountainous regions with environmental factors using geospatial data. It delivers an interactive dashboard and web interface that enable users to quickly query and visualize information by season and region, enhancing hiking safety awareness.

Workflow:

• Problem: In a one‑month Dataathon, our four‑member team—assembled on the same day after a workshop—faced limited TBIA open data, inconsistent citizen‑science records, and missing geocoordinates, yet needed to build an actionable dashboard in one week(2 weeks of prep, 1 week of build).
• Action & Result:
  • Led data pipelines in Python (Pandas, NumPy, requests, pyinaturalist, datetime) to clean and standardize data—filtering to 500+ high‑quality records—and deployed an interactive Tableau dashboard in under 7 days;
  • Earning 1st place out of 12 teams and a showcased our poster at the Citizen Science Carnival.

Objective:
This study analyzes the fluctuations of weekly influenza and complication case counts in Taiwan and builds forecasting models to evaluate their predictive performance. It finds that outpatient visits, emergency visits, and death counts exhibit clear seasonality, while severe local complication cases show intermittent patterns.

Workflow:

• Problem: In 2020, while global attention centered on COVID-19, influenza and its complications still ranked among Taiwan’s top 10 causes of death; we needed to understand its seasonal dynamics under pandemic conditions and build a reliable short-term forecasting model.
• Action & Result:
  • Collected and preprocessed 655 weeks of CDC flu data—applying log transforms, differencing, and stationarity tests—then implemented ARIMA/SARIMA, Croston’s, and Holt-Winters models in Python.
  • Achieved short-term forecast accuracy with MAPE < 10% (as measured by MAD and RMSE) by the pipeline, providing insights to inform public health decisions.

• AI Foundations & Knowledge Concepts: AI Overview; Big Data Advanced Concepts; Knowledge Representation
• AI Technical Concepts: Image Recognition; Speech Recognition; Machine Learning; Deep Learning
• AI Systems & Industry Applications: Smart City & Smart Home; Smart Medical Care & Public Health; Intelligent Education; New Retail & Customer Services; Intelligent Manufacturing; AI & Society Development

• Proficient in data-oriented programming: data structures & objects, relational/NoSQL databases, data import/export, functions & control flow, debugging & performance tuning.
• Skilled in data processing & analysis: data cleaning & summarization, feature transformation & extraction, big data concepts, probability & statistics fundamentals, exploratory data analysis (EDA), supervised & unsupervised learning.

• Exam covered descriptive statistics, probability distributions, hypothesis testing, and related topics.
• Awarded “Outstanding Award” for achieving a score of 90+ out of 100.