2023
- Triangulation and Desity Estimation based DBSCAN Clutering Algorithm
- We proposed a spatial clustering algorithm for nonlinear data with various density and touching issue (including adjacent problem, neck problem, etc).
- Accepted by 2023 Symposium of Statistics and Data Science to present as a speaker and host poster session
- Currently working on developing the algorithm into a python package
- MCM: Analysis of Wordle’s score distribution based on word attributes
- Create a word difficulty model by training word attributes data using Locally Weighted Regression, Support Vector Regression, and Decision Tree; use GridSearchCV to tune hyperparameters
- Use the word difficulty model to predict the number of score reported in the future and how the percentage of players in hard mode changed based on different word attributes